Development and spatial distribution of soils on an active volcano:

Mt Etna, Sicily

Peter James

⁎

, David K. Chester

a,b

, Angus M. Duncan

Department of Geography and Planning, School of Environmental Sciences, University of Liverpool, Liverpool, L69 3BX, UK

Department of Geography and Environmental Studies, Liverpool Hope University, Hope Park, Liverpool L16 9JD, UK

abstractarticle info

Article history:

Received 11 June 2015

Received in revised form 21 September 2015

Accepted 30 September 2015

Available online 22 October 2015

Keywords:

Mt Etna

Soils

Pedogenesis

Chronosequence

Lava

Tephra

As a large, active volcano, Mt Etna presents a complex soil-forming environment, with a spatial diversity of lava-

ﬂow age, tephra deposition, local climate and agriculture. Following a review of previous published research on

Etna's soils, this paper analyses physical and chemical data for 23 proﬁles between 120 m and 1030 m elevation,

on all ﬂanks of the volcano, with the aim of characterising the soils and determining the extent to which the chief

controls on their nature and geography may be identiﬁed. Soils are sampled across a chronosequence of lava

ﬂows, dating from 72 yr to between 28 ka and 42 ka BP, which occur in conditions of diverse land use and

local climate. Ages of ﬂows N 3000 yr are only broadly constrained. The majority of soils comprise mostly tephra.

Chief soil processes inferred have been identiﬁed in previous published research on Mt Etna: accumulation of soil

organic matter, rapid loss of Na and production of ﬁne-grained materials, particularly those inferred to be of

short-range order. Other developments are soil structural formation and evidence of increasing soil mesofaunal

activity. Despite the complexity of environmental diversity and spatial variation in tephra deposition, a number

of soil properties (including: material b.020 mm; organic C; organic N; C/N; pH; total Na; extractable phases of Fe

and Al) are statistically related to age of lava ﬂow, most of these also related to annual rainfall. That a number of

statistically signiﬁcant relationships with age of ﬂow also hold for near ground-surface samples is unexpected

and unexplained. Two areas which require further research are the precise impacts of tephra deposition on de-

velopment of Etnean soils, and processes of weathering and related pedogenesis at the soil–lava bedrock

interface.

1. Introduction and aims

This paper reviews previous research into the soils of the largest ter-

restrial volcano on the European mainland, the active Mt Etna (Fig. 1),

and analyses new data. Because of their global distribution, the human

settlements and economies they support, their distinctive properties

and signiﬁcance as an element of the volcanic environment, volcanic

soils have attracted much attention from specialists in pedology, ecolo-

gy, hydrology, environmental chemistry, horticulture and archaeology.

Soils may also contribute to the impact of volcanoes on the atmosphere:

rapid weathering of volcanic materials, a part of pedogenesis, may con-

stitute an important sink in the global CO

cycle (Vitousek et al., 1997).

Distinguishing between release of soil-derived CO

and the higher

degassing rates of magma-derived CO

is an important aid in predicting

volcanic activity (Liuzzo et al., 2013). Soil reactive processes moderate

storage and release to groundwater of volcanigenic elements potentially

harmful to humans (Bellomo et al., 2007; Floor et al., 2011).

The genesis of volcanic soils is reasonably well understood (for a

summary of chief processes, see Supplementary Appendix 1). Speciﬁc

process pathways, development rates and geography of soils on a par-

ticular volca no, however, are not readily predicted. Large, active

volca noes, especially those of predominantly Hawaiian/Strombolian

eruptive style such as Mt Etna, present some of the most complex soil-

forming environments on Earth. Historic activity on Mt Etna has been

predominantly Strombolian in style, but periodic phases of la va

fountaining of Hawaiian style generate considerable lapilli fall on the

ﬂanks. In the historic period there have been major explosive eruptions

including the plinian event of 122 BC which deposited up to 15 cm of

tephra on Catania some 3 0 km from the vent (Coltelli et al., 1998).

Several subplinian eruptions have also occurred (Branca and Del Carlo,

2005). Diversity in age and character of volcanic materials, land surface

morphology, local climate, vegetation and land use history translates

into complex soil spatial patterns determined by controls operating at

scales of differing spatial order, e.g. local climate, age and morphology

of lava ﬂow, and even individual plants (Certini et al., 2001). In the ver-

tical dimension (i.e. proﬁle), complex soils result fr om intermittent

tephra deposition, anthropogenic disturbance, erosion and subsequent

deposition. Soil proﬁles may reﬂect the amount and frequency of tephra

deposition as much as ‘normal’ proﬁle-forming soil processes operating

Catena 137 (2016) 277–297

⁎ Corresponding author.

E-mail addresses: jg04@liverpool.ac.uk (P. James), jg54@liverpool.ac.uk (D.K. Chester),

aduncan@liverpool.ac.uk (A.M. Duncan).

http://dx.doi.org/10.1016/j.catena.2015.09.023

Contents lists available at ScienceDirect

Catena

journal homepage: www.elsevier.com/locate/catena

on stable surfaces. The building-up of soil by the addition of tephra may

have either a ‘retardant’ or ‘developmental’ effect on pedogenesis

(Johnson, 1985). In the former case existing soil is effectively sealed

by burial, and pedogenesis begins anew in fresh material; in the latter,

pedogenesis keeps pace with a lesser amount of added tephra, absorb-

ing the new material, the soil nevertheles s being altered and made

more difﬁcult to interpret. A problem in dealing with the impact of

tephra deposition is its considerable spatial va riation, because the

Fig. 1. Mt Etna. a. Relief, geology and location of sampling sites. ‘Other units’ include basal tholeiites and pyroclastic deposits. b. Mean annual rainfall and wind direction (after Durbin and

Henderson-Sellers, 1981); chief settlements and place-names referred to in text. c. Map of Fierotti et al.'s (1988) soil associations. d. Location of Mt Etna and study area.

278 P. James et al. / Catena 137 (2016) 277–297

history of deposition at a particular location is unlikely to be known.

Furthermore, units of tephra in soil are not everywhere easily differen-

tiated in the ﬁeld.

Many analyses of soil spatial variation on volcanoes relate soil prop-

erties to environmental gradients which represent important state fac-

tors in pedogenesis. The gradients are sequences of age, climate (as it

varies with elevation and slope orientation), vegetation or topography.

Many volcanic environments are well suited to chronosequence studies

because of the age-range of lava surfaces. Examples abound. Vitousek

et al. (1997) applied the chronosequence approach to the study of soil

and ecosystem development across a well-deﬁned 1.4 m yr sequence

in Hawaii. Shorter age-ranges have been examined in Italy: e.g. on Mt

Amiata by Lulli (2007), and in three studies comprising the most com-

prehen sive published analyses of soil proﬁles on Mt Etna to date, by

Egli et al. (2007, 2008, 2012). Drawbacks of soil chronosequence model-

ling include the role of factors other than time and the problem of cli-

mate change over longer timescales (Huggett, 1998). There is also the

problem of determining age of landforms and/or sediments. Age esti-

mates for older (N 3000 ka) lava ﬂows on Mt Etna are only broadly

constrained (De Beni et al., 2011). Where appropriate and with careful

interpretation, however, chronosequence studies provide at least a

good basis to begin analysis of processes and patterns of soil develop-

ment across an area.

Ped ological research on Mt Etna has tended to co ncen trate either

on selected soil properties or processes, or on relationships between

soils and certain vegetation types or plant spec ies, and deals in most

cases w ith comp rehensive analyses of a small numb er of soil proﬁles.

This approach is essential to understanding the soils and the envi-

ronmental controls upon them. Of equal importance, however, are

the soils' spatial relationships with environmental controls as they

vary across the volcano. There are two small-scale soil maps of Mt

Etna (Comitato per la Carta dei Suoli, 1966; Fierotti et al., 1988),

but without accompanying published details on the soils. Egli et al.

(2007, 20 08, 2012) ﬁnd that a number of soil properties are related

to age of the substrate lava and to elevation (and therefore precipita-

tion, temperature and vegetation). In the present study we cast the

spatial net more widely, examini ng a sample of soils on all ﬂanks of

the volcano, and across a range of environments, inc luding those dis-

turbed by agriculture, and across a range of local climates. The soil at-

tributes selected for analysis are fundamental to soils generally and

to volcanic soils particularly. Our aim is to determine whether, and

to what exten t, environmental controls on pedogenesis and soil spa-

tial distrib ution are identiﬁable from the complex of inﬂuences af-

fecting the soils sampled. This is attempted using statistical tests

and by considering spati al variation in soil characteristics. Soil vari-

ables are ﬁrst plotted against age of lava ﬂow. The ﬂows comprise a

chronosequence. The soils in a strict sense do not: they have devel-

oped thr ough a range of past environmental conditions. Age of ﬂow

is, nevertheless, a sound basis for soil sampling, and the question of

the extent to which soil properties reﬂect ﬂow-age remains one to

be addressed.

Soils are described using the terminol ogy of Schoeneberger et al.

(2012). They are classiﬁed only tentatively, and as far as the derived

data permit, according to the Soil Taxonomy system (Soil Survey Staff,

1999).

Most Soil Taxonomy terms used in the text are deﬁned in the

glossary provided in the Supplementary Appendix 2. For further details

and for terms not included in the glossary, the reader is referred to the

Soil Survey Staff (1999, 2014).

2. Mt Etna

Volcanism in the Etnean region started during the middle Pleisto-

cene, at about 600 ka (Branc a et al., 2004). The vo lcano now spans

3350 m in elevation and 40 km in basal diameter; tephra fall has been

frequent. Below 1100 m above sea level (a.s.l.), the sampling zone (i.e.

120–1030 m), lava varies in age from the 2014/15 ﬂow to the ca

500,000 yr-old tholeitic basalts of a small area on the lower part of the

southern ﬂank (Gillot et al., 1994; De Beni et al ., 2011). Flows of the

last millennium (F ig. 1a), referred to below as ‘historical’, are dated

using written records; older (‘prehistoric’) ﬂows by archaeomagnetic

and radiometric, particularly U series, dating (Condomines et al., 1982;

Tanguy et al., 2007, 2012). Terrain of historical and of some prehistoric

ﬂows is dominated by rugged aa lava, pahoehoe and toothpaste

morphology

(Kilburn and Guest, 1993; Polacci and Papale, 1999;

Guest et al., 2012). The mineralogy and geochemistry of historical and

young prehistoric Mt Etna lavas vary little (Romano and Guest, 1979:

see Table 1; Tanguy et al., 1997; Corsaro and Pompilio, 2004). The

hawaiite lavas are porphyritic with up to 45% phenocrysts comprising

plagioglase, clinopyroxene and olivine (typically in that order of abun-

dance) set in a matrix of microlites and glass with scattered

titanomagnetite granules (Chester et al., 1985). Explosive activity from

summit craters is frequent. Less frequent eruptions, often with higher

effusion rates, occur from ﬂank vents, with Strombolian activity from

vents high on the volcano (Chester et al., 1985). The geochemistry of

the tephra, emitted from summit craters and lateral vents, is similar to

that of the lava (Coltelli et al., 2000). The tephra (which consists of phe-

nocrysts set in a glassy matrix) varies in deposition rate and particle-

size with distance and direction from source, and accumulates unevenly

on rugged lava surfaces. The detailed tephra chronostratigraphy at any

location is thus likely to be unique and unknown. Heiken and Wohletz

(1985) reported the media n particle-size of tephra samples from a

1971 eruption to vary between .001 and 1 mm. There are few tephra de-

posits comprising extensive stratigraphic markers (Del Carlo and

Branca, 1998). In the soils we examine, only one buried tephra layer is

ident i ﬁed, that of the 1669 eruption of the Monti Rossi cinder cones

(Fig. 1b). Tephra at our site ES19, on the 1792 ﬂow, is probably from a

recent eruption.

A Mediterranean climate of marked seasonality is recorded at a

number of stations on Mt Etna (Carpenter, 2004). Mean annual rainfall

(m.a.r.) increases with elevation and is signiﬁcantly greater on the E

than that on t h e W ﬂank (Fig. 1b):

Adrano, W ﬂank, 589 m. a.s.l.: m.a.r. 497 mm; January means: 69 mm,

9.8 °C; July m eans: 14.3 mm, 26 °C.

Zafferana, E ﬂank, 574 m a.s.l.: m.a.r. 1353 mm; January means:

227 mm, 9.2 °C; July means: 17 mm (June: 12.7 mm), 24.8 °C.

A strong prevalence of westerly to northwesterly winds (Fig. 1b;

Durbin and Henderson-Sellers, 1981)issigniﬁcant for the distribution

of airborne volcanic emissions.

Judging from the meteorological data and high permeability of

many soils, soil moisture regimes for the study area are xeric and

ustic; and temperature regimes predominantly thermic, with mesic

at the higher elev ations (see the Supplementary Appe ndix 2 for def-

initions). Fernández-Sanjurjo et al. (2003) judge the regimes at

Monti Rossi (850 m a.s.l., Fig. 1a) to be ustic and mesic. Geobotanical

regions of Mt Etna range with elevation through thermo, meso, supra

and oromedi terranean (Poli Marchese and Patti, 2000), the sampling

area for the present study lying within the ﬁrst two.

Below 900 m elevation the more productive soils are terraced for a

wide range of tree crops, vines and to a more limited extent for arable

crops (Dazzi, 2 007). Some terraces, especially at higher elevations, have

not been cultivated for a numb er of years, t hough these are s till grazed.

aa is lava where the surface crust of active lava ﬂow is disrupted into irregular frag-

ments of clinker that are rafted on the mobile interior of the ﬂow. Surfaces comprise blocks

in an openwork structure. Pahoehoe typically has smooth, undulating surfaces, sometimes

lobate or rucked up forming a ropy texture. Toothpaste is transitional between aa and

pahoehoe. It has a viscosity higher than that of pahoehoe and rate of ﬂow slower than aa.

As its name implies, it has a morphology that resembles toothpaste extruded from a tube.

279P. James et al. / Catena 137 (2016) 277–297

Woodland is common on older ﬂowsbetween900and2000ma.s.l.,acli-

mate cont rol evident in a preference of evergreen oak (Quercus i lex)for

the W side and deciduous oak (Quercus pubescens; Quercus cerris)and

economically impor tant chestnut (Castanea sativa) woods for the E side

of the mountain (Poli Marchese and Patti, 2000; Puzzolo and Folving,

2001). The E tna broom (Genista aetnensis), a common colonizer of open

areas, is found up to 1800 m a.s.l. Major ash fall may cause signiﬁcant

damage to fruit and vegetable crops ( Dazzi, 2007). Within our sampling

zone, broad relationships between vegetation on the one hand and age,

morphology a nd slope aspec t of lava ﬂow on the other, are v isi ble in the

landscape (as reﬂected in the site data provided in Supplementary

Table 1). At the sites examined, the Etna broom occurs on ﬂows older

than that of 1923; 3 m-high deciduous oak (Quercus sp.) on the 1865

ﬂow; 3.5 m everg reen oak (Q. ilex) on the 17 92 ﬂow; mature deciduous

oak woodland on the 1408, and mixed deci duous woodl and on prehist or-

ic ﬂows. Agricultural terraces or walled enclosures occur at ﬁve of the sites

(≥671 yr- old), and intense grazing at two (Supplementary T able 1).

Using palynological data from Lago di Pergusa in central Sicily

(667 m a.s.l., 70 km to the west of Mt Etna), Sadori and Narcisi (2001)

infer that the transition from dry Pleistocene to more moist Holocene

climate started at about 10,700 yr BP. The wettest conditions of the Post-

glacial occurred between ~9000–~7200 yr BP, followed by a trend to-

wards desiccation at about 3000 yr BP. Signiﬁcant human impact upon

vegetation and soils of Mt Etna is likely to have begun early, during

the Bronze Age, if not in Neolithic times (Leighton, 1999; Chester

et al., 2010).

In a few of the many papers published on the chemistry of gas and

particle emissions, rainwater and groundwater, implications are consid-

ered for rock weathering but not for pedogenesis. Degassing from the

summit craters continuously releases volatiles into the atmosphere,

with high enrichment of at least 13 trace metals together with Al, Fe,

and SO

in the volcani c plume (Buat-Ménard and Arnold, 1978;

Varekamp et al., 1986; Francis et al., 1998; Gauthier and le Cloarec,

1998; Varrica et al., 2000; Aiuppa et al., 2003a, 2003b). Mt Etna is the

world's greatest point source of F, SO

and HCl (Allard, 1997; Francis

et al., 1998), b 1% being deposited around the volcano during eruptions

(Aiuppa et al., 2006). Rapid downward mixing of the plume on to the

lower ﬂanks of the volcano, however, occurs under certain meteorolog-

ical conditions (Allen et al., 2006). Particulate emissions also contribute

signiﬁcantly to rainwater chemistry (Cimino and Toscano, 1998),

deposition rates of major ions being controlled mainly by rainfall

(Aiuppa et al., 2006). Na, K, Ca and Mg have varied sources, including

volcanic ash and aerosols. Volcanic emissions comprise a signiﬁcant ad-

dition to dust sources of many chemical elements in lichens on Mt Etna

(Varrica et al., 2000). Vacca et al (2003) and Egli et al. (2008) conclude,

however, that an allochthonous loessic component is of minor impor-

tance in volcanic soils of southern Italy. Passive degassing on Mt Etna's

ﬂanks results in intense leaching of basalt by phreatic groundwaters

rich in CO

(Giammanco et al., 1996; Aiuppa et al., 2000; 2004;

Alparone et al., 2004). The residence time of any gas-charged waters

in permeable soil is short, but the effect of magma-derived CO

on Mt

Etna soils has not to our knowledge been researched.

3. The soils of Mt Etna: review of research

Part of Fierotti et al.'s (1988) map is reproduced in Fig. 1c. In this,

soils are classiﬁed according to Soil Survey Staff (1975).Associations1

and 10, largely at higher elevations, were classed as Lithic, Typic and

Andic Xerorthents, and Andic Xerochrepts; Association 15 at lower ele-

vations, and extensive on E, S and SW ﬂanks, included Typic

Xerorthents, Andic Xerochrepts and Ultic Haploxeralfs. In the scheme

of Soil Survey Staff (1975) a separate Andisol Order had not been intro-

duced (they were in 1978); the soils classed by Fierotti et al. as ‘andic’

would probably include soils now classed as Andisols. The Ultic

Haploxeralfs, the most fully developed of the soils mapped, are referred

to as clayey and as havin g A–Bt–Cproﬁle which includes well struc-

tured, organic-rich, dark brown or black A horizon and blocky, more

often prismatic, argillic B horizon. Dazzi (2007) and Lo Papa et al.

(2003) (both referred to by Lulli, 2007), encountered only young, vitric

soils on Mt Etna, Lo Papa ﬁnding Typic Vitrixerands, Typic Udivitrands

and Vitric Hapludands, all under forest (and all Andisols).

Published papers analyse pedological aspects (Bellanca et al., 1996;

Buemi et a l., 1998; Busà et al., 1998; Certini et al., 2001 and Parisi et al.,

2003); effects of plants on soils (Certini et al., 2001; Fernández-Sanjurjo

et al., 2003) and biogeochemical cycling (Leonardi and Rapp, 1981;

Leonardi et al., 1992; Leonardo et al., 1996; Santa Regina et al., 2001). Min-

eralization of carbon in soils of Mt Etna was analysed by Hopkins and

Bartoli (2004), and soil microbial activity by Hopkins et al. (2007) and

Shillam et a l. (2008). Plant colonisation and lichen-weathering of lavas

have been examined by Carpenter (2004).

In rece nt studies on or near the cinder cones on the south ﬂank of

the volcano, soils classiﬁe d by variou s authors include: sand-skeletal,

mixed, mesic, aridic Ustipsamment (beneath broom on tephra of

Monti Rossi, 850 m a.s.l., Fig. 1c); sand-skeletal, mixed, frigid

Vitrandic Udorthent (under broom and pine – Pinus nigra – on

Monte Vetore, 1820 m a.s.l., Fig. 1c) and Vi tric Hapludands (elevation

not given; precipitation, 1130 mm yr

− 1

)(Bellanca et al., 1996; Busà

et al., 1998; Del Carlo and Branca, 1998; Palumbo et al., 2000; Certini

et al., 2001; Fernández-Sanjurjo et al., 2003). The weakly weathered

Vitric Hapludand reported by Palumbo et al. (2000) contains allo-

phane, imogolite and halloysite. Andic properties do not occur in

the Monti Rossi and Mt Vetore soils described by Fernández-

Sanjurjo et al. (2003),butwereconﬁrmed by Bel lanca et al. (1996)

in a soil in which allophane, imogolite, halloysite and a small amount

of goethite were evident in the B horizon. The former authors de-

scribe different characteristics between soil under the canopy of

bro om and that outside the canopy on Monti Rossi. The sand fraction

of these soils is lar gely vesiculated gl ass, with plagioclase, augite,

magnetite and apatite. Organic matter and NH

oxalate-extractable

Al, Si and Fe increased, and Ca, Mg, K and Si de creased down-

proﬁle, reﬂecting greater alteration of tephra in the B than in the A

horizon formed in overlying younger tephra. Busà et al. (1998) also

describe a reduction in Ca, Mg and K, and relative enrichment in Al,

Fe and Ti concentrations at 1372 m a.s.l. near the Monte Salto del

Cane cone (Fig. 1c). Contrasts between soil proﬁles on Mt Vetore

(1800 m a.s.l.) and Monti Rossi (850 m a.s.l.) are attributed by

Fernández-Sanjurjo et al. (2003) to differences in climate and more

intense pedogenesis associated with plant roots at the lower site.

Certini et al. (2001) conside r the greater acid weathering under

pine than under Etna broom on Mt Vetore to be a cause for concern

regarding loss of soil fertility, as t he Corsican pine (P. nigra)has

been widely planted on Mt Etna in recent deca des. Those authors at-

tribute low amounts of N in the Vitrandic Udorthent of Mt Vetore in

part to coarse mineral texture and continual soi l rejuvenation by

fresh cinder. Total C, microbial biomass and microbial respiration

were analysed by Hopkins et al. (2007) in eight soils under trees be-

tween 850 m and 1 450 m a.s.l. Carbon content and microbial biom ass

were g reater in more developed soils, but ability to analyse trends

with soil development was limited by differences in management

Table 1

Selected oxides (percent weight) for Etna lavas, dating from 1923, 1947, 1809 and 1614–24.

Romano and Guest (1979).

Min Max

SiO

46.33 48.86

17.35 18.94

5.48 7.06

FeO 3.01 5.82

MgO 4.44 5.85

CaO 9.53 10.37

O 3.32 4.40

O 1.50 1.76

280 P. James et al. / Catena 137 (2016) 277–297

and tephra addi tion. Shillam et al. (2008) compared the stru ctural

and functional properties of the microbial communities in two soils

of different ages on Mt Etna. The microbial community in the older

soil was structurally more diverse than t hat of the younger soil but

proved in experiments not to be more resistant to environmental

disturbance simulated by experimental heating.

The important role of reactive surfaces of ﬁne materials of Mt Etna

soils on the storage and release of volcanigenic pollutants to groundwa-

ter has been demonstrated. Bellomo et al. (2007) estimate that soils ab-

sorb ~60% of atmospheric F depositions, maintaining F within the safe

drinking-water limit; the ac tive soil constituents being short-range

order minerals (SROMs), particularly allophane. Both SROMs and soil

organic matter are implicated by Floor et al. (2011) for Se immobiliza-

tion, but with competition by sulphate for adsorption sites and mineral

dissolution by acid rainwater (on the E ﬂank) accounting for release of

adsorbed Se to groundwater (with less Se mobilized in organic-rich

soils).

Egli et al. (2008) examined eight soils (6 on the N, 2 on the SE ﬂank)

between 780 and 1090 m a.s.l., on lavas 125 to 115,000 yr old and with

estimated mean annual rainfalls ranging between 1100 and 1200 mm.

Only soils on non-agricultural land were sampled. The soils, classiﬁed

using IUSS Working Group WRB (2006), were equivalent to Entisol on

the 125-yr old lava, Inceptisols on 358 and 438 yr BP lavas and Andisols

on 115,000 yr BP lavas. The main mineral transformations were inferred

to be volcanic glass → imogolite -type minerals → kaolinite. Textures

were predominantly sand and loamy sand. These authors found the

chief pedogenic processes to be leaching of base cations and passive en-

richm ent of Al and Fe, and that these were pronounced in the early

stages of pedogenesis, with much reduced rate of cha nge between

15,000 and 115,000 yrs. In contrast, the rate of formation of imogolite-

type minerals and particularly of ferrihydrite was more linear through

the chronosequence. Imogolite was transformed to kaolinite, amounts

of 2:1-type clay minerals being very small even on the oldest ﬂows,

and inferred to have been released from glass by weathering rather

than resulting from pedogenic transformation. The scarcity of the 2:1

type clay minerals was attributed to addition of ash and low rainfall.

The accumulation rate of soil organic matter was most rapid in ﬂows

up to 2000 yr old, tending to zero after 15,000 yr. Stabilisation of organic

matter was related to pyrophosphate-extractable Al and Fe phases, ka-

olinite and clay content, rather than to imogolite-type material.

Baglieri et al. (2010) ﬁnd, however, that glycerol-extractable soil organ-

ic fractions were closely bound to SROMS, particularly in older

(N 9000 yr) soils (Udivitrands), and consequently less subject to biodeg-

radation. From their ﬁrst analysis of ﬁve Vitric Andosols (IUSS Working

Group WRB, 2006) in an elevation sequence (551 to 1772 m) on 10,000

to 15,000 yr BP lava, on the north-east ﬂank of Mt Etna, Egli et al. (2007)

conclude that organic C stocks, (and imogolite-type minerals and crys-

talline Fe-oxyhydroxides) decreased with elevation.

In a study of their elevation sequence of soils, Egli et al. (2012) ad-

dress the complex and important issue of ﬁre impacts on the soils. The

stable soil organic fraction at higher elevations survived from

N 8000 yr, dating close to the onset of pedogenesis, whereas with de-

creasing elevation soil charcoal analysis indicates a history of increased

ﬁre frequency causing vegetation change and rejuvenation of soil or-

ganic matter. The authors conclude, however, that ﬁre frequency on

Mt Etna is not such as to deplete a large pool of soil organic matter. In

a subse quent s tudy of this elevation sequence, Ma

strolonardo et al.

(2013) illustrate the importance of ﬁre on the quality and quantity of

soil organic matter on Mt Etna. Fire frequency is inferred to be a factor

in stocks of organic C and N and of C resistant to chemical oxidation

(all increasing with decreasing elevation). These authors stress the

major importance of vegetation and local climate as factors in amounts

and nature of the organic matter in the soils they analyse.

Bellanca et al. (1996) account for differences between A an d B hori-

zons in one soil to an age-difference of tephra between the horizons. In

the present study we ﬁnd boundaries between buried tephra units

difﬁcult or impossible to identify, and no morphologically distinct bur-

ied soil was found. Buried units of tephra may be deﬁned by detailed

grain size and chemical analyses of soil proﬁles (Taboada et al., 2007),

but there are few data on buried soils in tephra sequences of Mt Etna

(Zehetner et al., 2003). Superimposed buried soils (classiﬁed as Typic

Hapludands) in two deep exposures of radiocarbon-dated tephra

sequences are described by Agnelli et al. (2007). These, at 1350 m a.s.l.

(S ﬂank) and 1550 m a.s.l. (E ﬂank), are now under chestnut

(Castanea sativa) and pine forest (Pinus laricio), respectively. The base

of the oldest sequence sampled dates from ~7000 yr BP, the timescales

for developmen t of the buried soils being much shorter than this,

b 800 yr in one case. Neoformed ‘amorphous’ and crystalline mineralog-

ical phases were absent, a feature attributed to insufﬁciently intense, or

too short a period, of pedogenesis.

4. Materials and methods

Soils were sampled at 23 sites in 1999. Sites (Fig. 1a) cover a range in

age of subjacent lava ﬂow (72 yr to ~35,000 yr), elevation, vegetation

and agriculture. The soils examined occur in teph ra (of varying

particle-sizes) deposited on lava and accumulating particularly in cracks

and depressions, though in two soils alluvial deposition of local soil ma-

terial was evident. Below, soils are referred to by site number with age

of ﬂow in brackets. The ages of lava ﬂows (Supplementary Table 1)

are based on the Geological Map of Etna Volcano ( Branca et al., 2011).

The dating of lavas older than reliable historic records (N~400 yr BP)

is tentative and where possible based on archaeomagnetic and radio-

metric dating (Tanguy et al., 2012), though dates for some older ﬂows

are constrained only by stratigraphical relationships. For the statistical

analysis of the soils the age of ﬂow for each site is recalibrated against

a baseline of 2000 AD. For those lavas constrained only by a range of

age (in some cases, very broad), a date at the mid-point of the range is

used (Supplementary Table 1). Archaeological evidence (Drs. Gianna

Ayala and Francesco Privitera, pers. comm.) was used in the ﬁeld

where possible. Where surface and margins of older

ﬂo

ws are concealed

by well developed plant cover and where agricultural use is intensive it

may be difﬁcult to identify ﬂows in the ﬁeld, or to know the age of the

underlying lava. ES23 posed a particular problem, this being a dagala

of unknown age (Supplementary Table 1). Samples from this site are

therefore excluded from all statistical analyses which involve the age

of ﬂow.

Annual rainfall for study sites (Supplementary Table 1) is estimated

by interpolation from data quoted by Durbin and Henderson-Sellers

(1981) (Fig. 1b). Sites and soils were described in the ﬁeld, following

Schoeneberger et al. (2012). Where feasible, soil pits were dug to bed-

rock or to clast-supported lava blocks (horizons denoted by ‘R’ in Sup-

plementary Table 1). As tephra stratiﬁcation was impossible to deﬁne

in most cases (c.f. Buurman et al., 2004), the horizon notation used in

Supplementary Table 1 gives the pedological character of each horizon

but does not deﬁne sequences of buried layers, which inevitably occur

(Agnelli et al., 2007). As we do not have all data required for full classi-

ﬁcation following Soil Taxonom y (Soil Survey Staff, 1999), we name

soils tentatively at the Order or Suborder level only. The three criteria

for Andisol classiﬁcation not determined are P adsorption, bulk density

and volume proportion of volcanic glass (Soil Survey Staff, 1999). The

high glass content observed in many relatively young soils on Mt Etna

(which we classify as Entisols) will reduce the threshold amounts of

SROMs required to qualify a soil as A ndi sol (Soil S u rvey Staff, 2014 ), but

on this criterion alone, ES1(1928 AD); ES2(1928 AD); ES3(1928 AD);

ES9(1809 AD); ES12 (1651 AD) and ES22(1843 A D) would fail to qual ify

as Andisols (year of emplacement of lava ﬂow in parentheses after site

number).

Prior to laboratory analysis, samples were air-dried, disaggregated

and passed through a 2 mm sieve. Sub-samples for particle-size analysis

were not dried further in order to avoid irreversible aggregation on dry-

ing where amorphous materials are present (Nanzyo et al., 1993). Laser

281P. James et al. / Catena 137 (2016) 277–297

diffraction was used for particle-size determination of b 1mmmaterial

in preference to the sieve and pipette method as it was considered pru-

dent to keep samples in suspension for as short a time as possible. Tex-

tural classes named in Supplementary Table 1 are based on estimation

in the ﬁeld (following Mizota and van Reeuwijk, 1989). We infer the na-

ture of neoformed materials from chemical data rather than from direct

physical examination. For phases of Al, Fe and Si conventional subscripts

are used, as shown in Table 2. The conventional procedures were used

for extraction of : 1. organic-bound (by pyrophosphate), denoted

below as e.g. Fe

; 2. inorganic non-crystalline (by NH

oxalate), denoted

as e.g. Fe

o-p

, and 3. crystalline phases of Fe and Al (by dithionite), denot-

ed as e.g. Fe

d-o

. These data are to be treated with caution as, for example,

oxalate may dissolve a portion of the crystalline phase (García-Rodeja

et al., 2007; Agnelli et al., 2007; Algoe et al., 2012).

In the following sections, after discussion of proﬁle morphology, the

analytical variables are examined in relation to age of ﬂow and location

in terms of ﬂank of the volcano, the latter varying signiﬁcantly in rainfall

amount and tephra fallout from summit eruptions. For brevity, the

moister ﬂanks of the volcano (i.e. NE, E, SE and S) are referred to as E

ﬂank, other sectors as W ﬂank. Spearman correlation (ρ =coefﬁcient;

p = probability of null hypothesis) and the Mann–Whitney U test are

used. For parametric statistical analyses non-normal variables are trans-

formed using log

, square root, and in one case (Fe

d-o

) an inverse trans-

formation. Of the functions tested for regression models relating

transformed soil variables to age of ﬂow (Figs. 2 and 10a) the quadratic

proved efﬁcient in terms of signiﬁcance level in the great majority of

cases . Multivariate analyses used are principal components analysis

(pca) and cluster analysis. In the former, orthogonal components are ro-

tated using Varimax. The unsupervised, agglomerative, hierarchical

cluster analysis using the ﬁrst four components of the pca is based on

the Euclidean distance as the dissimilarity measure and Ward's group-

ing method, with all variables standardized to mean of zero and vari-

ance of 1.

5. Results

5.1. Proﬁle morphology

Details of proﬁle morphology and site conditions are given in Sup-

plementary Table 1. Table 3 groups soils according to extent of develop-

ment expressed in proﬁle morphology. On this criterion, there is a clear

trend of increasing pedogenesis with age of ﬂow between the groups, I

to IV (Table 3). Only in the two soils of group IV, ES6 (15–3.9 ka) and

ES13 (450 AD), was there a clear B horizon and a colour other than

black (with colour value ≥ 3). Very shallow soils, Psam ments (I,

Table 3), occur in rock-dominant areas of (mostly young) lava ﬂows

with very thin (commonly b 5 cm) accumulations of tephra and other

allochthonous mineral and organic material in depressions in bedrock

surfaces or between blocks in aa lava. Vegetation is limited to lichens

or mosses, or may also include a narrow range of vascular plants. Exam-

ples are ES1, ES2 and ES3 on the 1928 Mascali ﬂow. Organic matter con-

tent may be relatively high under a more developed plant cover

(Supplementary Table 1). The soils are black (2.5Y 2.5/1), sandy, cindery

or ashy, vitric, and occur in pockets (typically b 30cm in diameter) in ex-

posed lava where organic debris and small amounts of tephra collect.

Commonly, ash and lapilli are trapped within the plant cover. On aa

ﬂows, mosses bridge narrow interstices between the lava clinker.

Smooth, little weathered surfaces of pa hoehoe lava on a 1991 ﬂow

collect mineral and plant detritus, moisture and rabbit dung in depres-

sions, and support a number of plant species and larvae. The lava sur-

faces beneath these soils had no sign of weathering visible in the ﬁeld.

Vingiani et al. (2013) report that mineral fragments which have been

transported to accumulate in cavities and depressions on young lava

ﬂows of Mt Etna are little altered chemically beneath a cover of lichens.

Developed in thicker tephra than those soils described above, are

coarse-textured Entisols (or Andisols) with limited visible evidence of

pedogenesis (II in Table 3). Sufﬁcient amounts of glass, value of

(Al

+½Fe

)% and horizon thickness (≥ 18 cm) would qualify some of

these soils as Vitrandic Xeropsamments or Vitrandic Ustipsamments

(depending on moisture regime). Tephra is of highly variable thickness

(7 cm to N 120 cm), depending in part on the underlying lava morphol-

ogy. With the exception of reddish brown Monti Rossi (Red Mountains)

tephra in ES20 (1329 AD) the soils are black (predominantly 2.5Y 2.5/1

and 10YR 2/1), ashy, cindery or pumiceous, sandy, occasionally skeletal,

loose, stony or very stony, non-sticky and non-plastic. Such soils are

very common in the s tudy area on ﬂows of all morphologies, dating

from 1923 AD to 1329 AD in the cases we sampled. They have O, A or

Ap horizons between 3 and 15 cm thick over little altered ash, lapilli,

cinder, clinker, pumice or scoria. The soils vary in texture, horizon thick-

ness and depth to lithic contact. Otherwise, variation in proﬁle mo rphol-

ogy is limited to very weak, barely percep tible structure in some A

horizons. Organisms observed in these soils in February were larvae,

millipedes and ants. Roots commonly extend to bedrock. Vegetation in-

cludes herbs and grasses, often with mosses and, in some cases, lichens.

Broom is c ommon (to 4 m in height); deciduous oaks ar e 3 m h igh near

ES14 (1865 AD), 8 m near ES15 (1689 AD) and mature near ES21

(1408 AD); evergreen oak grows to 6 m at ES12 (1651 AD) and ES20

(1329 AD). Land has been organized for terraced agriculture only at

ES20 (1329 A D). Here, dark redd ish-brown (2.5YR 3/4) c ind er from the

1669 Monti Rossi ﬂank eruption at the base of the soil appears to have

been used to level the lava surface for terrace construction. There was ev-

idence of grazing by sheep at ES9 (1809 AD) and by rabbits at other sites.

Other soils developed in tephra and lacking a B horizon are Entisols

or Andisols (III in Table 3)ofﬁner texture than in groups I and II,

Table 2

Soil properties determined, techniques used, and units expressed. Except for particle-size fractions, b 2 mm material is analysed.

Soil properties Unit Method and description

Total Ca, Mg, K, Na mg g

−1

Digestion in nitric and perchloric acids (Hesse, 1971); determination of Ca and Mg by atomic absorption spectrometer; of

Na and K by colorimeter using molybdenum blue method (Allen, 1989).

Particle size fractions

(in b 1 mm material)

% vol Coulter laser diffractometer, following pretreatment with H

, and dispersion with ‘Calgon’. Textural terms based on

ﬁeld estimation because of the unreliability of particle-size analysis of Andisols (Nanzyo et al., 1993).

Pyrophosphate-extractable Fe and Al

(Fe

,Al

)

mg g

−1

Extracted with potassium pyrophosphate (McKeague, 1967). For Fe and Al present in metal–organic complexes.

Oxalate-extractable Fe, Al and Si

(Fe

o-p

,Al

o-p

,Si

o-p

)

mg g

−1

Extracted in darkness with acid NH

-oxalate (McKeague and Day, 1966). Inorganic, non-crystalline.

DCB-extractable Fe (Fe

d-o

)mgg

−1

Extracted with dithionite–citrate–bicarbonate (Mehra & Jackso n, 1960). Crystalline Fe oxyhydroxides (goeth ite, magnetite, etc.)

pH in water − log [H+] Electrometrically in water (2.5:1 volume water:soil)

Organic carbon % weight Potassium dichromate digestion by the Walkley–Black method (Hesse, 1971)

Organic nitrogen % weight Kjeldahl procedure (Allen, 1989)

Fig. 2. Soil variables plotted against log of age (x axis) of lava ﬂow for all samples. Black circles: W ﬂank sites; open circles: E ﬂank. Curves for quadratic (or linear) function and 95% limits

shown in each case. R

=coefﬁcient of determination. For samples with equal or very nearly equal values, symbols are over-written (e.g. samples in ES17 and ES18 in several cases).

282 P. James et al. / Catena 137 (2016) 277–297

283P. James et al. / Catena 137 (2016) 277–297

containing loamy sand or sandy loam. Age of ﬂow varies from 1180 ±

30 AD (ES7) to 42–28 ka (ES17 and ES18). Where exposed, lava mor-

phology includes slabby aa, pahoehoe, levé e and tumulus.

The soils

are black, with A or Ap and C horizons. The ﬁner texture of these soils

results from either a ﬁner parent tephra (ES4 15–3.9 ka) or from more

advanced weathering than in those soils described above. In the silty

clay loam of ES17 and ES18 (both in the range, 42 to 28 ka), no material

coarser than sand-size was observed. Related to the ﬁner texture is a

weak granular or blocky structure in A horizons of some soils and the

presence of earthworms (Supplementary Table 1), which, where more

common, cause marked bioturbation: in ES17, earthworm burrows, ex-

tending to the base of the soil, are a striking feature. Consistence varies

from non- to moderately sticky, but slightly friable in ES8 (15–3.9 ka).

All sites have a mixed herb and grass ﬂora; a number have been terraced

for agriculture, recent cultivation having been abandoned in several

cases. Trees are deciduous oak: 3 m high at ES7 (1180 ± 30 AD) and

6matES8(15–3.9 ka); olives (Olea eu ropaea)atES10(15–3.9 ka)

and olives and prickly pear (Opuntia ﬁcus-indica) in a walled enclosure

at ES4 (15–3.9 ka). In all cases where lava surfaces were examined be-

neath soils they were ‘clean’, irrespective of age of ﬂow. Red tephra of

the Monti Rossi 1669 eruption occurred in ES24 (1000 ± 50 AD) (Sup-

plementary Table 1).

The most fully developed soils examined have a Bw horizon (IV,

Table 3). These are either Inceptisols or Andisols, are very dark greyish

brown (10YR 3/2) to brown (10YR 4/3), silt loam over clay loam (ES6

15–3.9 ka) and loam (ES13 1651 AD). There was no evidence of recent

tephra fall at these sites, both of which occur on the W ﬂank of the vol-

cano. A horizons are blocky; the Bw in ES13 being blocky and in ES6,

prismatic. ES6 is distinct in most respects from other soils examined.

Unusually for Etnean soils, the lower pa rt of the proﬁle is not free-

draining, slight gleying occurring on the ped faces of the lower part of

the Bw horizon. The soil, at least in part, appears to have been washed

into the depression where the soil occurs, which may account for the

presence of pottery fragments at a 33–34 cm depth. ES6 and ES13 are

in enclosed ﬁelds and are the most intensively farmed soils examined,

ES6 being grazed by cattle and ES13 cultivated. Trees are 4 m-high de-

ciduous oak, almond (Prunus dulcis) and prickly pear at ES6, and pista-

chio (Pistacia vera) at ES13. There is a trend of increasing pedogenesis

with age of lava ﬂow, as reﬂectedinsoilproﬁle morphology (with the

exception of ES6 and ES13), from I to IV. Soils are listed by age but with-

in each group it is not possible to separate soils on the basis of extent of

development.

5.2. Soil physical and chemical properties

Analytical data for all samples are presented in Supplementary

Table 2. Values for selected variables (for each soil sample) are plotted

against age of ﬂow in Fig. 2, in which the E and W ﬂanks (as deﬁ

ned

in Se

ction 4) are differentiated. Where available estimates give a

broad range of age, the mid-point of the range is used (Section 4). In

Table 4a are shown correlations for variables with age and rainfall, re-

spectively; also listed (Table 4b) are variables which differ signiﬁcantly

(using the Mann–Whitney U test) between the W and E ﬂanks of the

volcano, with the ﬂank noted on which the median of the variable is

greater.

5.2.1. Particle-size

Soil texture, as estimated in the ﬁeld, ranges from coarse sand to clay

loam. The size distribution and proportion of particles N 2mmvaryap-

preciably (Supplementary Table 1), tending to be of high concentration.

Fine particle-size fractions correlate strongly with age of ﬂow, as illus-

trated for b.002 mm (clay) in Fig. 2a and by the envelopes of particle-

Table 3

Summary of general soil and site characteristics grouped into (tentative) soil Orders (I to

IV). There is a trend of increasing pedogenesis with age of lava ﬂow (with the exception of

ES6 and ES13), from I to IV. Soils are listed by age but within each group it is not possible to

separate soils on the basis of extent of development.

Soil Age of

ﬂow, yr

Broad soil charact eristics

ES1 72 I Very shallow, black Psamments or Orthents in pockets in

lava. Plant cover predominantly mosses and lichens; few

vascular plants.

ES2 72

ES3 72

ES11 77 II Black soils, including Entisols, Psamments (possibly

Andisols) in little-altered tephra. Soil depth and range of

soil mesofauna are greater than in group I, though no

evidence of earthworms was found. Occasionally very weak

structure in A horizons. Highly variable vegetation from

bryophytes and grasses to mature trees. Some agricultural

terracing.

ES14 135

ES22 157

ES9 191

ES19 208

ES15 311

ES12 349

ES21 592

ES20 671

ES7 820 III Black loamy Entisols, possibly Andisols; occasional weak

granular or blocky structure; earthworms. Vegetation

includes mature cropped trees associated with terraced

agriculture.

ES24 1000

ES5 1050

ES23 Unknown

ES8 ~9500

ES4 ~9500

ES10 ~9500

ES17 ~35,000

ES18 ~35,000

ES13 1550 IV Greyish-brown to brown, loamy and ﬁne loamy Inceptisols;

possibly Andisols: blocky A (over prismatic Bw horizon in

ES6). ES6 is pasture; ES13 pistachio orchard.

ES6 ~9500

Table 4a

Spearman correlations with age of ﬂow and rainf all for all samples and near-surface samples

only (only signiﬁcant correlations shown). Correlations signiﬁcant at ≥95% in bold. pc =

principal component score (Table 5). ES23 (dagala within 1780 AD ﬂow) is omitted from

analyses involving age of ﬂow.

All samples Near-surface samples only

Age Rain Age

b.002 mm ρ .760 −.591 .868

p b.001 .000 .000

.002–b.02 mm ρ .748 −.561 .871

p b.001 b.001 b.001

Organic C ρ .272 −.017

p .015 .879

Organic N ρ .445 −.271

p b.001 .012

C/N ρ −.455 .593 −.422

p b.001 b.001 .050

pH ρ .324 −.320

p .003 .003

Ca ρ −.156 .388

p .166 b.001

Mg ρ −.165 .258 −.424

p .143 .018 .049

Na ρ −.850 .379 −.929

p b.001 b.001 b.001

K ρ .135 −.601

232 b.001

ρ .711 −.404 .531

p b.001 b.001 .011

o-p

ρ .275 .256

p .013 .019

d-o

ρ .719 −.607 .853

p b.001 b.001 b.001

o-p

/Fe

d-o

ρ −.286 .675 −.509

p .010 b.001 .016

ρ .607 −.489 .598

p b.001 b.001 .003

o-p

ρ .314 −.272 .449

p .005 .012 .036

o-p

ρ −.170 .107

p .131 .331

pc1 ρ .766 −.449 .862

p b.001 b.001 b.001

pc4 ρ − 247

p .027

pc5 ρ .256

p .022

284 P. James et al. / Catena 137 (2016) 277–297

size curves in Fig. 3. Silt, .002–b.02 mm, also correlates strongly with age

(Table 4a). The b.002 mm fraction correlates negatively with rainfall

(Table 4a), there being a signiﬁcant difference in b.002 mm material be-

tween the E and W ﬂanks (Table 4a), the lowest concentrations occur-

ring across the SE and S ﬂanks (Fig. 4). Despite intermittent addition

of tephra to many of the soils, for near-surface samples only, the corre-

lation between b.002 mm and age of ﬂow is higher (ρ = .868; p b .001)

than that for all samples (Table 4a). For the deepest samples taken in

each soil, the correlation with age of ﬂow is ρ =.781,pb .001. The sig-

niﬁcant correlation with age of ﬂow for near-surface samples obtains for

the variables listed in Table 4a, the degree of correlation being high in a

number of cases.

As allochthonous particles have been discounted as a signiﬁcant

component of the soils of Mt Etna (Section 2), we attribute the increase

in the ﬁner fractions to pedogenesis. The map of proﬁles of b.002 mm for

each soil (Fig. 4)reﬂects age-dependency and factors related to location

on the volcano. Intense weathering in organic-rich, near-surface hori-

zons is a likely cause of an up-proﬁle increase in b.002 mm material

(Fig. 4) in a number of ﬂows up to 1050 yr BP (ES5 950 ± 50 AD; ES9

1809 AD; ES15 1689 AD; ES21 1408 AD; ES22 1843 AD and ES24

1000 ± 50 AD), but more marked is a down-proﬁle increase in older

ﬂows (partic ularly in ES6 15–3.9 ka; ES13 1651 AD; ES17 42–28 ka;

ES18 42–28 ka) resulting from either of, or both, weathering of older

tephra at depth and weathering of lava at the lithic contact. Argillic ho-

rizons (Bt) which formed as a result of illuviation of clay are common in

well developed soils of the Mediterranean region (Ya

alon, 1997). Soils

with Bt horizons are mapped at lower elevations of Mt Etna by Fierotti

et al. (1988) (Fig. 1c), and they occur in older volcanic soils of southern

Italy (Lulli, 2007). We have not undertaken micromorphological analy-

ses, but ﬁnd no evidence indicating clay illuviation in the soils we

sample.

5.2.2. Organic carbon

The marked near-surface enhancement of organic matter in many of

the soils (Fig. 5) is the most visible morphological effect of pedogenesis

across the full age-range of ﬂows sampled, and on all ﬂanks of the vol-

cano. Organic C concentration is nowhere of sufﬁcient depth to form a

melanic epipedon. Charcoal may be a cause of black coloration in volca-

nic soils, but in the soils we examine the colour is determined by little-

weathered, bas altic parent mater ial. Owing to the sensitivity of soil

organic matter concentration to plant cover, ‘dilution’ by tephra

and other disturbance, amoun ts of organic C vary widely. Hi gh con-

centrations (N 8.0%) occur in surface horizons beneath a range of

vegetation types: cult ivated fruit tr ees (ES 4 15–3.9 ka: 8.96%);

mature deciduous oak (ES21 1408 AD: 9.8%); discontinuous herb

and grass cover in a lava levée (ES5 950 ± 50 AD: 8.3%), and beneath

the lichen of ES2 (1928 AD: 8 .8%). There is no distinct regional

pattern in organic C proﬁles (Fig. 5), and organic C does not correlate

with rainfall (Table 4 a). There is, however, a weak but signi ﬁcant

correlation be tween organic C and age of ﬂow (ρ = .263, p = .016;

Fig. 2b). In their sample of soils Egli et al. (2008) ﬁnd a steady

increase of soil organic C stock to an age of about 15,000 yr.

Hopkins et al. (2007) also ﬁnd that more developed soils have higher

concentrations of organic C (and N) on Mt Etna.

Table 4b

Signiﬁcant differences (95% level) between W and E ﬂanks, using Mann–Whitney U.

Signiﬁcant differences

U Flank with greater median

d-o

1354.5 W

o-p

/Fe

d-o

169 E

o-p

456 E

1157.5 W

1171 W

o-p

600.5 E

C/N 328 E

.02–1000 mm 339 E

.002- b 20 mm 1185 W

b.002 mm 1256 W

b.001 mm 1241 W

K 1249 W

Ca 503 E

Na 461 E

Mg 576 E

pH 1065.5 W

pc1 1162 W

pc3 386 E

pc4 521 E

Not signiﬁcant

Organic C

Organic N (p = .53)

o-p

pc2

Fig. 3. Envelopes of cumulative particle–size curves (.001–1 mm) for soils on ﬂows

grouped by age: a. 72 to 208 yr BP: ES1, ES2, ES3, ES11, ES14, ES22, ES9 and ES19. b. 311

to 1050 yr BP: ES15, ES12, ES21, ES20, ES7, ES24 and ES5. c. ~9000 to ~35,000 yr BP:

ES4, ES6, ES8, ES10, ES13, ES17 and ES18. ES23 — age unknown.

285P. James et al. / Catena 137 (2016) 277–297

Fluctuations in organic matter content with depth would be expect-

ed in soils which have received intermittent tephra deposits or have

been disturbed in some other way. In the majority of soils sampled

there is, however, a progressive decline in organic matter content

with depth (Fig. 5). In the moderate concentrat ions of organic C at

depth in the agricultural terrace soil, ES8 (15–3.9 ka) (4.47% at 23 cm

depth, and 5.59% at 53 cm), there is a suggestion of soil burial resulting

from terrace construction.

The quality of organic matter in terms of organic N and the ratio, C/N,

is more closely related to age of ﬂow than is organic C (evident in the

correlations, Table 4a). The possible inﬂuence of rainfall amounts on or-

ganic matter quality, though not organic C, is also evident in correlations

with organic N and C/N (Table 4a), and also in the signiﬁcant E–W ﬂank

differences in C/N but not in organic C (Table 4a). Organic N just fails the

95% level for signiﬁcance of this difference. Lower C/N ratios in Andisols

of drier xeric as compared with udic moisture regimes may be attribut-

ed to a higher rate of mineralization in the former (cf Broquen et al.,

2005).

5.2.3. pH

pH varies between 4.8 and 7.7, many values being higher than those

reported by Shoji et al. (1993) for volcanic soils in other climatic regions,

higher than reported by Certini et al. (2006) for the largely trachydacitic

Mt Amiata, Tuscany, but comparable with values found by Vacca et al.

(2003) on Roccamonﬁna and by Egli et al. (2008) on Mt Etna (5.1 to

6.25). Most of the values b 5.5 occur on the E ﬂank (Fig. 6), presumably

reﬂecting acid deposition from the plume and higher rainfall. There is an

increase in pH with age (ρ .307, p .004) (Fig. 2c; Table 4a), which is more

marked with ES17 and ES18 (both 42–28 ka) samples omitted for this

correlation: ρ =.610andp=b.001. It is not clear why this increase oc-

curs. The correlation between pH and rainfall is negative, and there is a

signiﬁcant pH difference between the E and W ﬂanks (Table 4b). There

are, however, variations on each ﬂank with, for example, anomalously

low values which are difﬁcult to explain in ES17 and ES18 (42–28 ka)

(Fig. 2

c), samples ES17/1 to 3 yielding the only values b 5.

5.2.4. Bases

In common with ﬁndings of Egli et al. (2008), Na is the most respon-

sive of the bases to weathering and leaching. In the present study, of all

those measured, this property shows one of the most marked changes

with age of ﬂow (correlation: ρ − .850, p b .001), the rate of loss show-

ing no sign of abating over the ~35,000-yr range (Fig. 2d). The highest

values for Na (reaching 7.27 g kg

− 1

) occur in the Psammen ts, 11

(1923 AD) and ES22 (1843 AD) (Supplementary Fig. 1). The signiﬁcant

correlation with age for near-surface samples only is very high and neg-

ative for Na (Table 4a). At variance with the Na trend is the lack of

Fig. 4. Schematic map of clay: b.002 mm % against depth for each soil. Exact location of sample sites shown in Fig. 1a.

286 P. James et al. / Catena 137 (2016) 277–297

correlation between age of ﬂow and Ca and Mg, and lack of at least

linear correlation with age for K (Figs . 2e–g; Table 4a). This contrasts

with the ﬁndings of Egli et al. (2008),thatbasecationleachingisa

dominant process in their sample of Mt Etna so ils, and of Martínez

Cortizas et al. (2007), who report a reduction in K and Ca (as well

as in Na), with an increasing degree of weathering in a wide range

of European volcanic soils. Such element depletion Martínez

Cortizas et al. (2003) attribute to the diluting effect of organic mat ter

on element concent ration, and the mobility of certain elements.

Nevertheless, they ﬁnd that, despite sig niﬁcant depletion of Ca and

K in soils of the Gauro (Campi Flegrei) and Vico volcanoes, Ca en-

hancement occurs in organic-rich horizons, which they attribute to

biocycling of the element. In most soils we analyse, Ca and more es-

pecially K show an up-proﬁle increase (Supplementary Figs. 2 and 3),

these cat ions correlating negatively with d epth in the soil (Ca: ρ −

.364, p .001; K: ρ −.312, p .004). Ca correlates positively with organic

C(ρ .396, p b .001). Quality of organic matter (and therefore possibly

its age) appears, however, to have greater control on base distribu-

tion than organic C concentration: K correlates negatively with the

C/N ratio. Ca, Mg and Na (in contrast t o its negative correlation

with organic C) correlate positively with C/N ( correlation details

not shown). A factor in greater near-surface Ca, Mg and Na concen-

trations in a number of soils we sample may be soil-rejuvenation

by tephra. A signiﬁcant distinction o f Mt Etna among most of its

European volcanic peers is that it is continually active .

Na, Mg, Ca and K all correlate with rainfall (K negatively: Table 4a),

and all show signiﬁcant differences between the E and W ﬂanks, the

ﬁrst three tending to occur in g reater amount on the E ﬂank (Table 4b).

The E–W contrast may reﬂect greater tephra fall on the E ﬂank. This, how-

ever, wo uld not a ppear to b e a major factor in K distribution.

5.2.5. Extractable Fe

Concentrations of extractable Fe phases – organic-bound (Fe

), non-

crystalline (Fe

o-p

) and cr ystalline (Fe

d-o

) – are plotted for e ach sample in

Fig. 7. For the majority of samples, the relative abundance of the three

phases is in the order, Fe

o-p

N Fe

d-o

N Fe

, a feature of t he Etnean soils re-

ported by Fernández-Sanjurjo et al. (2003),andonereﬂecting both the

status of the SROM, ferrihydrite, and the limited pe dogenesis in m uch of

the study area. T he stronger corr elations with age of ﬂow, however, a re

for the crystalline (Fe

d-o

: ρ .719, p b .001) and organic-complexed (Fe

ρ .711, p b . 001) phases, bot h showing a steepening increase o n ﬂows of

N 200 yr (Fig. 2 h–j). The ratio, Fe

/Fe

, related to the crystallinity of the

Fe oxides (Schwertmann, 1985 ), has b een found to reﬂect the degree of

pedogenesis in volcanic soils, the ratio value decreasing with age

(Malucelli et al., 1999). This trend i s indicated in our samples by a wea k

but signiﬁcant negative correlation between Fe

o-p

/Fe

d-o

and age of ﬂow

Fig. 5. Schematic map of % organic carbon against depth for each soil. Exact location of sample sites shown in Fig. 1a.

287P. James et al. / Catena 137 (2016) 277–297

(Table 4a). In contrast, on volcanic materials of Roccamonﬁna Volcano

(older t han those w e analyse) Vacca et al. ( 2003) ﬁnd Fe

/Fe

not t o be

age-dependent. In our M t Etna samples, Fe c rystallisation is favoured by

seasonal dehydration in the Mediterranean climate (Schwertmann,

1985), which is drier on the W ﬂank, where median F e

d-o

is signiﬁcantly

greater than on the E (Table 4b). The correlation between Fe

o-p

/Fe

d-o

and rainfall is quite strong, and t here is a signiﬁcant difference between

the E and W ﬂanks (Table 4b). Fe

o-p

/Fe

d-o

shows no consistent depth pat-

tern between soils, but a clear down-proﬁle increase in several of the soils

(Supplementary Fig. 5: ES7 1180 ± 30 AD; ES19 1792 AD; ES23 age un-

known; ES21 1408 AD on the E ﬂank, and ES12 1651 AD and ES17

35,000 bp on the W), the proportion of crystalline secondary Fe reﬂecting

more advanced transformation in layers of older tephra, and/or less crys-

tallinity i n organic-rich ho rizons.

5.2.6. Extractable Al and Si

The presence of Al and Si in the form of SROMS is inferred from the

range of Al

o-p

(.26% to 2.44%) and Si

o-p

(.064% to 1.36%), the ﬁrst corre-

lating with age of ﬂow, but not the latter (Table 4a; Figs. 2l and n). Lack

of correlation between each of these phases and organic C is not consis-

tent with the protection of organic matter by SROMS, commonly in-

ferred for volcanic soils, though degree of humiﬁcation m ay be a

factor: Al

o-p

correlates negatively with C/N (ρ −.315, p .004). As Fe

the organic-bound Al

increases with age, the Al and Fe organic phases

being quite strongly inter-correlated (ρ .787, p b .001). In most samples,

pH (N 5.0) is favourable to allophane and imogolite formation, and al-

though Al

increases with age, Al

/Al

o-p

is b 0.5 (a condition associated

with the presence of allophane and imogolite: Legros, 2013)inallsam-

ples except ES6/2 to 4 (15–3.9 ka) and 13/3 (450 AD). Al

and Al

/Al

o-p

correlate with organic C (Al

: ρ .449, p b .001; Al

/Al

o-p

: ρ .447, p b .001),

but both have negative correlations with C/N (Al

ρ −.432, p b .001;

/Al

o-p

: ρ −.315 p .004), reﬂecting the ability of the humiﬁed fraction

to compete for Al, thereby reducing Al availability for the formation of

allophane.

The value of the qualifying criterion for Andisols, (Al

+½Fe

)%, in-

creases with age of ﬂow (Fig. 2o), a number of samples in soils on

N 9000-yr ﬂows having values N 2.0 (samples ES6/4 15–3.9 ka; ES8/2

and 3 15–3.9 ka; ES17/1, 2 and 3 42–28 ka), but also the much younger

ES14 (1865 AD) (samples 3 and 4). However, high glass content in-

creases the potential for qualifying as Andisols on this criterion, a

value of (Al

+½Fe

)% = 0.4% being the minimum required (Soil

Survey Staff, 2014, p. 18). This value is exceeded in all samples analysed,

though in many, other qualifying criteria for Andisols which are not

measured are unlikely to be met.

Bellanca et al. (1996) found that NH

oxalate-extractable Al, Fe and

Si concentrations all increase down-proﬁle in soils to the north of

Fig. 6. Schematic map of pH (H

O) against depth for each soil. Exact location of sample sites shown in Fig. 1a.

288 P. James et al. / Catena 137 (2016) 277–297

Nicolosi. In our greater ra nge of soils, there is no consistent down-

proﬁle trend in Al

o-p

or Si

o-p

(Figs. 8 and 9). The reason for higher

amounts of Si

o-p

occurring on the N side of the volcano is not clear.

6. Principal components analysis

The variables selected for the pca and the rotated component matrix

are shown in Table 5.Theﬁrst four components (pcs) with initial eigen-

values N 1 account for only 70% of the variance but give a meaningful in-

sight into combinations of soil variables reﬂecting key dimensions of

pedogenesis on the volcano. Pc1 includes those properties shown to

be related to age of ﬂow and which reﬂect, and may be used as one in-

dicator of, the extent of pedogenesis as regards transformations involv-

ing Fe, Al and Si, the four bases analysed, organic matter and particle-

size distribution. High loadings obtain for measures of ﬁne particle-

size, Na (negative), Fe

d-o

(inverse positive) and organic C. The highest

negative loading for C/N ratio in the matrix measures the greater degree

of humiﬁcation associated with this component. The high loading for

relative to the negative loading of Si

o-p

gives a non-allophanic signal

in the component, and reﬂects the increase of Al

with age of ﬂow.

Sample scores on pc1 correlate with age of ﬂow and negatively with

rainfall (Fig. 10a; Table 4a), there being a signiﬁcant difference in scores

between the volcano's E and W ﬂanks (Table 4b), giving a marked re-

gionalization to scores; negative scores occurring on the SE and S sides

of the volcano (Fig. 10b), possibly reﬂecting greater tephra addition on

these slopes. An up-proﬁle increase in scores occurs in ES5 (950 ±

50 AD); ES12 (1651 AD) ; ES22 (1843 AD); ES4 (15–3.9 ka); ES9

(1809 AD); ES11 (1923 AD); ES7 (1180 ± 30 AD), ES15 (1689 AD);

ES23 (age unknown), ES20 (1329 AD); ES21 (1408 AD) and ES24

(1000 ± 50 AD) (Fig. 10b), indicating greater alteration, as deﬁned by

this component, in near-surface horizons. This is contrary to the ﬁnd-

ings of Bellanca et al. (1996) regarding more intensive alteration in B

than in A horizons, referred to in Section 3.Adown-proﬁle increase in

scores occurs , however, in four of the most developed soils (ES18

35 ka; ES13 450 AD; ES6 15–3.9ka;ES815–3.9 ka).

In pc2, the two phases, Si

o-p

and Al

o-p

, are isolated with their high

scores consistent with an allophanic effect. The component lacks indica-

tors of advanced pedogenesis with loadings low for ﬁne particle-sizes

and positive for Na. With Al

/Al

o-p

inserted into the analysis in lieu of

and Al

o-p

(result not shown), the structure of components 1 and 2

is little changed, but the loading of Al

/Al

o-p

is high positive in pc1 and

moderate negative in pc2, reinforcing the distinction between condi-

tions associated with an allophanic versus a non-allophanic effect. Pc2

correlates signiﬁcantly with neither age of ﬂow nor rainfall, nor is

there a signiﬁcant difference in scores between E and W ﬂanks. There

is no clear regionalization in pc2 scores. A down-proﬁle increase in

scores occurs (not shown) in eleven soils: ES18 (42–28 ka); ES4 (15–

3.9 ka); ES6 (15–3.9 ka); ES8 (15–3.9 ka); ES 22 (1843 AD); ES14

(1865 AD); ES19 (1792 AD); ES20 (1329 AD); ES21 (1408 AD); ES23

(age unknown) and ES24 (1000 ± 50 AD), which may relate to decreas-

ing organic matter content with depth.

pc3 is a moderate Fe

o-p

and positive (inverse) Fe

d-o

component, the

resulting ratio between these phases suggesting the role of ferrihydrite.

The component does not correlate with age, but positively with rainfall

(ρ .523, p b .001: Table 4a), and there are signiﬁcantly greater scores on

the E ﬂank (Table 4b), consistent with conditions favouring persistence

of a high Fe

o-p

/Fe

d-o

ratio. The less organic-rich soil at depth may explain

a predominant down-proﬁle decrease in scores (not shown), occurring

in all soils except ES6 (15–3.9 ka), ES12 (1651 AD), and ES22 (1843 AD).

A feature of pc3 scores not shared by pc1 or pc2 is a reduction in rate of

decrease down-proﬁle, or a reversal in trend, in the lowest parts of the

sampled proﬁles at twelve locations, ES4 (15–3.9 ka); ES5 (950 ±

50 AD); ES6 (15–3.9 ka); ES8 (15–3. 9 ka); ES9 (1809 AD); ES12

(1651 AD); ES14 (1865 AD); ES19 (1792 AD); ES20 (1329 AD); ES21

(1408 AD); ES23 (age unknown) and ES24 (1000 ± 50 AD), though

scores remain negative.

pc4, a ‘relatively high C/N–organic C–base cation’ component, is the

only component with positive loadings for all four bases, Ca, Mg, Na and

K. These are not associated in this component with signals of much ad-

vanced pedogenesis in terms of particle-size, Fe

o-p

/Fe

d-o

and C/N ratio.

Across the matrix, loadings differ between the bases , presumably

reﬂecting the interplay between parent material, loss from the soil dur-

ing pedogenesis and retention by plants, the most salient features being

the marked loss of Na and a putative retention of at least K by plants. pc4

scores do not correlate with age; they are generally higher on the E ﬂank

(Tabl e 4b), but there is no distinct regional pattern in scores (not

shown). A down-proﬁle decrease occurs in scores of all soils except

ES11 (1828 AD), reﬂecting the distribution of organic matter. The reduc-

tion in rate of decrease, or reversal of trend, at depth in the proﬁle found

in pc3 scores occurs also in scores on pc4, with the exception of ES10

(15–3.9 ka); ES11 (1923 AD); ES12 (1651 AD); ES15 (1689 AD) and

ES17 (42–28 ka). There is insufﬁcient evidence to speculate on the rea-

son for this common, if su btle, change in conditions at depth in the

proﬁle.

In pc5 a high positive loading for pH contrasts with much lo wer

loadings for other variables. The low % variance explained by this com-

ponent suggests a limited impact of pH in distinguishing between the

soils analysed, a reﬂec tion possibly of a narrow range of pH across

most of the samples. The low but positive loading for pH in pc1 reﬂects

the (unexplained) statistically signiﬁcant increase of pH with age of

ﬂow described above.

Fig. 7. ThreeFephasesforeachsoilsample,mgg

−1

:Fe

(black); Fe

o-p

(light grey) and Fe

d-o

(dark grey). Soils in order of increasing age of ﬂow. Position of ES23 – age unknown – is

arbitrary.

289P. James et al. / Catena 137 (2016) 277–297

7. Cluster analysis

Five clusters of samples are chosen from the analysis based on the

ﬁrst four pcs (Fig. 11), clusters 1 and 2 being a division of a group of

thirty-nine samples. The distribution of scores on the ﬁrst four pcs is

shown for each cluster in Fig. 12. The spatial distribution of sample

members in the ﬁve clusters (Fig. 13) comprises a form of soil map for

the sites sampled. In cluster 1 both median and lower quartile are pos-

itive on all ﬁve pcs, providing a pedogenic ‘mixed message’ in which sig-

nals of relatively advanced formation occur with expressions of a strong

presence of Al, Si and Fe SROMS (with highest mean of the clusters for

o-p

,Si

o-p

and Fe

o-p

: 10.8, 5.3 and 12.3 m g g

−1

respectively), and high

amounts of organic matter characterised by high C/N. Cluster 1 mem-

bers occur on all ﬂanks, especially in the NE, and in soils ranging from

135 yr to ~9500 in age (Fig. 13). In this cluster, the range in age and cli-

mate of sites, and the positive median score for all four pcs, give an im-

pression of a complex of controls on pedogenesis leading to relatively

similar outcome for all member samples.

Cluste r 2 (19 samples) is distinguished by the highest pc3 scores

(the ‘high Fe

o-p

/Fe

d-o

–high Mg’ pc) of all clusters. Sample scores on the

remaining pcs are largely negative. With the exception of ES5 (950 ±

50 AD), this cluster does not occur on the west ﬂank (Fig. 13), cluster

members oc curring at sites with the highest rainfall (i.e. 750 to

1250 mm ).

In Cluster 3 (20 samples), scores on pc1 and pc2 are largely negative.

The cluster is distinguished by the highest median on pc4 (the ‘high or-

ganic matter–high C/N–high base pc’). Samples occur on the E, SE and S

ﬂanks and in the younger soils of the W ﬂank (ES22—1843 AD and

ES12—1651 AD).

Cluster 4 (9 samples) has the lowest scores of all clusters on pc1, and

mainly positive scores on the ‘allophanic’ pc2. The cluster has the lowest

distribution of scores on pc3, (the ‘ferrihydrite’ pc). The weakly devel-

oped samples, which occur in ES11(1923 AD), ES22(1843 AD) and

ES9(1809 AD), comprise the coarsest tephra (highest .02–1.0 mm %;

lowest b.002 mm %) and the highest Na amount (mean: 5.2 mg kg

−1

)

of all clusters.

Cluster 5 (19 samples) is distinguished by its high positive median

on pc1 and its largely negative scores on other pcs. Unlike other samples

in this cluster, those in the upper 45 cm of ES17 (42–28 ka) have high

positive score on the ‘high Si

o-p

–Al

o-p

’ component, pc2. Samples in this

cluster, which express a relatively well developed character, occur on

the oldest ﬂows with the least rainfall, on the W ﬂank; though single

samples occur anomalously near the surface of ES4 (15–3.9 ka), ES5

(950 ± 50 AD) and ES9 (1809 AD).

Fig. 8. Schematic map of Al

o-p

:mgg

−1

against depth for each soil. Exact location of sample sites shown in Fig. 1a.

290 P. James et al. / Catena 137 (2016) 277–297

8. Discussion

8.1. Soil processes and proﬁles

The chief processes inferred from our data are largely those in-

ferred in a number of previous studies of Mt Etna, particularly that

of Egli et al. (2008). However, although loss of Na is one of the

most marked features of our sequence of soils, we do not, in contrast

to Egli et al. (2008) and Busà et al. (1998), have evidence of loss of Ca

or K. In our sample, the latter are associated with surface, organic-

rich hori zons and retained in the soil, a feature reﬂected in pc4,

which subsumes high amounts of organic matter and hig h C/ N ratios

together with positive loadings for all bases. Other chief processes

we infer are: production of clay- and silt-size material; incorporation

of organic matte r and its modiﬁcation in terms of decreasing C/N

ratio; the formation of SROMS (indic ated by NH

-extraction); devel-

opment of soil st ructure, and a tendency for vigorous bioturbation in

ﬁner g rained, older soils. We do not identify allophane directly, but

the pca clearly distinguishes conditions favouring allophane forma-

tion. A protective mechanism relating amou nts of organic C to NH

extractable phases is not borne out in the d ata. The degree of

humiﬁcation (C/N ratio) of organic mat ter, in a ddition to its

Fig. 9. Schematic map of Si

o-p

:mgg

−1

against depth for each soil. Exact location of sample sites shown in Fig. 1a.

Table 5

Principal components analysis: loadings of variables and % of variance explained. NB: in-

verse transformation used for Fe

d-o

Rotated component matrix

Component

12345

Log Fe

.907 −.124 −.014 .116 .007

Log .002–b.02 mm .904 .069 −.226 .062 −.010

Log Na −.896 .131 −.018 .243 .043

Log b.002 mm .895 .004 −.225 .149 .105

Sqrt Al

.805 .274 −.274 −.098 .072

Inverse Fe

d-o

−.752 −.184 .332 .197 −.258

o-p

.591 .042 .481 .249 −.359

Log Si

o-p

−.247 .891 .146 .019 −.095

Sq rt. Al

o-p

.398 .850 .020 −.170 −.065

K .202 −.005 −.853 .094 −.021

Mg −.203 .190 .626 .268 .082

Ca −.051 −.189 .226 .724 −.052

Sqrt org. C .553 −.005 −.053 .679 .046

C/N −.472 .146 −.121 .492 −.072

pH .114 −.123 .066 −.026 .955

% variance explained 37.7 11.8 11.6 10.34 7.70

291P. James et al. / Catena 137 (2016) 277–297

concentration, emerges as a signiﬁcant inﬂuenceonsoilchemistry.

An up-proﬁle increase in pc1 scores in many soils indicates greater

near-surface intensity in certain major processes.

Soils of limited development, generally with distinct organic-rich A

horizon in some cases overlain by an O horizon, but lacking B horizon,

occur on ﬂows dating up to 42–28 ka. These organic-rich horizons and

Fig. 10. a. pc1 scores plotted against log of age of lava ﬂow (x axis) for all samples. Black circles: west ﬂank sites; open circles: east ﬂank. Linear best-ﬁt and 95% limits shown. R

=co-

efﬁcient of determination. For samples which have equal or near-equal values, symbols are over-written (e.g. in ES17 and ES18). b. Schematic map of pc1 scores against depth for each

soil. Vertical line = score of 0 in each case. Exact location of sample sites shown in Fig. 1a.

292 P. James et al. / Catena 137 (2016) 277–297

a weak structure in the A horizon are the most distinct pedogenic ef-

fects. With increasing age of ﬂow, a ﬁner texture occurs and bioturba-

tion is more in evidence. Only two soils have Bw horizons, one (ES6

15–3.9 ka) formed in ﬂuv ially deposited volcanic material, the other

(ES13 450 AD), a rich soil in a mature agricultural landscape. We do

not verify the presence of illuviated clay using micromor phological

techniques, but we ﬁnd no suggestion of illuviation having taken

place. Nor are we able to quantify secondary crystalline clay minerals

in the soils.

In one agricultural terrace there is a buried soil (ES8 15–3.9 ka), and

two pr oﬁles (ES6 15–3.9 ka and ES17 15–3.9 ka) contain inwashed

sediment, but there is no clear evidence of major disruption of proﬁle

development in other soils, despite the identiﬁcation of buried tephra

units in two of the proﬁles. With the exception of Mg (Supplementary

Fig. 4), which we are not able to explain, many properties measured

do not have very irregular depth functions which would be expected

in soils comprising tephra layers of differing age. The near-surface en-

hancement of organic matter, which occurs in all but one soil, is typical

of proﬁles which have developed on relatively stable surfaces.

8.2. Environmental factors

Although the soils do not form a true chronosequence, their relation-

ship with age will be formed in part by the oldest material in each soil,

regardless of addition of fresh tephra. At least as important as ﬂow-age

in determining soil character and distribution on the volcano are rainfall

and other, unmeasured, environmental contrasts between windward

and lee sides of the volcano. The size of Mt Etna and direction of prevail-

ing wind have major implications for the volcano's soil geography. Mt

Etna is unique in its combination of volcanic activity and local climates.

On the E and SE ﬂanks, rainfall amounts should favour pedogenesis, yet

this appears to be restrained (as reﬂected in low pc1 scores), possibly by

tephra depos ition . A complex interpla y between effects of ﬂow-age,

tephra deposition and location on the volcano is reﬂected in the map

of cluster membership of samples (Fig. 13).

Lulli (2007) concludes that the Mediterranean climate of volcanic re-

gions of Italy does not favour allophane genesis and the full develop-

ment of Andosols, but conditions favourable to allophane formation

on Mt Etna are indicated in our analyses, and Andisols are classiﬁed in

a number of studies referred to in Section 3. Soils with illuviated clays

were mapped on Mt Etna by Fierotti et al. (1988), and a reddened B ho-

rizon was reported in a Typic Hapludand in late Pleistocene pyroclastic

material at 838 m a.s.l. on Vico Volcano, Italy (Quantin et al., 1985), but

in the soils we examine, even on the drier, more ‘Medite rranean’ W

ﬂank, and on the 42–28 ka ﬂow, neither red dening nor evidence of

clay illuviation was found. This contrasts with Pleistocene volcanic

soils elsewhere in the Mediterranean region: reddened soils with pedo-

genic calcite and illuviated clay in Pleistocene andesites of Methana,

Gree

ce, (m.a.r. b 400 mm; James et al., 1997) and on Quaternary basalts

of Morocco (Hamidi et al., 1999; Dekayir and El Maâtaoui, 2000); and

reddened soils in Pleistocene scoria of the Golan Heights, Israel (m.a.r.

850–950 mm: Singer et al., 2004).

8.3. Conclusion

We demonstrate the difﬁculties which may face the application of

chronosequence models to soil studies in complex volcanic environ-

ments. Our sample of soils is spread across a range of ecological, climatic

and agricultural environments. Ages of prehistoric lava ﬂows are only

broadly constrained. For all ﬂows, known or estimated age gives only

the date of onset of lava-weathering and related pedogenesis. At un-

known intervals, tephra has contributed soil parent material, this

forming the bulk of soils we examine. Two of the soils contain material

inferred to be washed from surrounding ﬂows. The sample of soils

therefore comprises only an approximate chronosequence. Soil variabil-

ity within single ﬂows is not examined explicitly, though, for example,

Fig. 11. Dendrogram from cluster analysis using sample scores on four principal compo-

nents. Cluster numbers 1 to 5 are as in Figs. 12 and 13.

293P. James et al. / Catena 137 (2016) 277–297

the contrast between ES17 and ES18 (on the 28–42 ka ﬂow) in terms

at least of sedimentology and soil thickness illustrates the signiﬁ-

cance of lava-ﬂow reli ef for so ils. We infer certain mineralogical

properties from chemical analyses, themselves arguably appr oxi-

mate. The classiﬁcation of soils is only partial, and, owing to the ex-

actitude of modern soil classiﬁc ation s chemes, wo uld requi re

determination of further soil properties to complete. Whatever the

ﬁner classes to which Mt Etna's soils may be allocated, however,

the majority of those we sample are entirely, and inevitably, ‘volca-

nic’ in their appearance and major properties.

The heuristic approach we take is chieﬂy statistical, with subjective

interpretation of spatial patterns of selected properties. Our sample of

soils is small, in view of Mt Etna's diverse environments. Nevertheless,

the number of signiﬁcant correlations, corroborated in many cases by

geographical distribution, appears to indicate a causal relationship be-

tween a number of soil properties and, respectively, age of lava ﬂow

and rainfall (and presumably other important environmental contrasts

between the volcano's ﬂanks). This ﬁnding echoes the conclusion of

Colombo et al. (2007) with respect to volcanic soils of south-central

Italy, that such factors as duration of pedogenesis, human impact and

soil climate appear to play more important roles than effects of stratiﬁed

pyroclastic materials. Not investigated in the present study is the impact

of vegetation on pedogenesis. The importance of plant cover is evident

from previous research into the soils of Mt Etna (Section 3) and is

reﬂected in the general relationship between vegetation and age of ﬂow

(Supplementary Table 1). Patterns we are not able to explain include

the increase of pH with age of ﬂow and the higher Si

o-p

and Al

o-p

concen-

trations o n the N slope.

That age of ﬂow emerges as a chief soil-forming factor in the com-

plexity of the Etn ean environment may be remarkable. Puzzling,

however, is the correlation between age of ﬂow and a number of

soil properties for near-surface samples only. This ﬁts easily wi th

neither ‘retardant’ nor ‘developmental’ effects (sensu Johnson,

1985) of intermittent tephra deposition on soil proﬁle developme nt.

The complexity of the effects of tephra addition remains a challenge

for research into the soils of Mt Etna. So too do processes at the base

of soils and the contribution of lava-bedrock weathering to soil pro-

ﬁles. The soil-rock interface remains a largely unexplored soil envi-

ronment on the volca no.

Supplementary data to this article can be found online at http://dx.

doi.org/10.1016/j.catena.2015.09.023.

Acknowledgements

The authors are grateful to the Royal Geographical Society (HSBC

Small Grant) for covering part of ﬁeldwork costs; to the Universities of

Fig. 12. Box plots showing median and upper and lower quartiles of scores on each of the four pcs for each cluster. Each pc is identiﬁed on the y axis.

294 P. James et al. / Catena 137 (2016) 277–297

Liverpool and Bedfordshire; to Sandra Mather for producing the illustra-

tions; to Alan Henderson, Irene Cooper and Bob Jude for the laboratory

analyses; to Drs. Gianna Ayala and Francesco Privitera for the archaeo-

logical data, and to the three referees for their constructive comments.

References

Agnelli, A.E., Corti, G., Agnelli, A., Del Carlo, P., Coltelli, M., Ugolini, F.C., 2007. Features of

some paleosols on the ﬂanks of Etna volcano (Italy) and their origin. Geoderma

142, 112–126. http://dx.doi.org/10.1016/j.geoderma.2007.08.003.

Aiuppa, A., Bellomo, S., Brusca, L., d'Alessandro, W., Fedrico, C., 2003a. Natural and anthro-

pogenic factors affecting groundwater quality of an active volcano (Mt Etna, Italy).

Appl. Geochemistry 18, 863–882. http://dx.doi.org/10.1016/S0883-2927(02)00182-

Aiuppa, A., Bonfanti, P., d'Alessandro, W., 2003b. Rainwater chemistry at Mt Etna (Italy):

natural and anthropogenic sources of major ions. J. Atmos. Chem. 46, 89–102. http://

dx.doi.org/10.1023/A:1024878323823.

Aiuppa, A., Allard, P., D'Alessandro, W., Giammanco, S., Parello, F., Valenza, M., 2004. Mag-

matic gas leakage at Mount Etna (Sicily, Italy): relationships with the volcano–tec-

tonic structures, the hydrological pattern and the eruptive activity. In: Bonaccorso,

A., Calvari, S., Coltelli, M., Del Negro, C., Falsaperla, S. (Eds.), Mount Etna: Volcano Lab-

oratory. Geophys. Monogr. Ser. 143. AGU, Washington, pp. 129–145 http://dx.doi.

org/10.1029/143GM09.

Aiuppa, A., Allard, P., d'Alessandro, W., Michel, A., Parello, F., Treuil, M., Valenza, M., 2000.

Mobility and ﬂux of major, minor and trace metals during basalt weathering and

groundwater transport at Mt Etna volcano (Sicily). Geochim. Cosmochim. Acta 64,

1827–1841. http://dx.doi.org/10.1016/S0016-7037(00)00345-8.

Aiuppa, A., Bellomo, S., Brusca, L., d'Alessandro, W., di Paola, Longo, M., 2006. Major-ion

bulk deposition around an active volcano (Mt Etna, Italy). Bull. Volcano l. 68,

255–265. http://dx.doi.org/10.1007/s00445-005-0005-x.

Algoe, C., Stoops, G., Vandenberghe, R.E., Van Ranst, E., 2012. Selective dissolution of Fe–Ti

oxides — extractable iron as a criterion for andic properties revisited. Catena 92,

49–54. http://dx.doi.org/10.1016/j.catena.2011.11.016.

Allard, P., 1997. Endogenous magma degassing and storage at Mount Etna. Geophys. Res.

Lett. 24, 2219–2222. http://dx.doi.org/10.1029/97GL02101.

Allen, S.E., 1989. Chemical Analysis of Ecological Materials. Blackwell Scientiﬁc,

Oxford.

Allen, A.G., Mather, T.A., McGonigle, A.J.S., Aiuppa, A., Delmelle, P., Davison, B., Bobrowski,

N., Oppenheimer, C., Pyle, D.M., Inguaggiato, S., 2006. Sources, size distribution, and

downwind grounding of aerosols from Mount Etna. J. Geophys. Res. - Atmos. 111.

http://dx.doi.org/10.1029/2005JD006015 (10 pp.).

Alparone, S., Andronico, D., Giammanco, S., Lodato, L., 2004. A multidisciplinary approach

to detect active pathways for magma migration and eruption at Mt Etna (Sicily, Italy)

before the 2001 and 2002–2003 eruptions. J. Volcanol. Geotherm. Res. 136, 121–140.

http://dx.doi.org/10.1016/j.jvolgeores.2004.05.014.

Baglieri, A., Ioppolo, A., Nègre, M., Abbate, C., Gennari, M., 2010. Characterization of glyc-

erol-extractable organic matter from Etna soils of different ages. Pedosphere 20,

120–128. http://dx.doi.org/10.1016/S1002-0160(09)60291-5.

Bellanca, A., Hauser, S., Neri, R., Palumbo, B., 1996. Mineralogy and geochemistry of soils

related to Etna volcano: heavy metal distribution as a result of soil-forming processes

and inherited features. Mineral. Petrogr. Acta 39, 95–107.

Bellomo, S., Aiuppa, A., D'Alessandro, W., Parello, F., 2007. Environmental impact of mag-

matic ﬂuo

rine emission in the Mt Etna area. J. Volcanol. Geotherm. Res. 165, 87–101.

http://dx.doi.org/10.1016/j.jvolgeores.2007.04.013.

Branca, S., Del Carlo, P., 2005. Types of eruptions of Etna volcano AD 1670–2003: implica-

tions for short-term eruptive behaviour. Bull. Volcanol. 67, 732–742. http://dx.doi.

org/10.1007/s00445-005-0412-z.

Fig. 13. Schematic map showing cluster membership of samples in each soil. Depth in each proﬁle not to scale. Clusters as deﬁned in Fig. 11. Exact location of sample sites shown in Fig. 1a.

295P. James et al. / Catena 137 (2016) 277–297

Branca, S., Coltelli, M., Groppelli, G., 2004. Geological evolution of Etna Volcano. In:

Bonaccorso, A., Calvari, S., Coltelli, M., Del Negro, C., Falsaperla, S. (Eds.), Mt Etna

Laboratory VolcanoGeophysical Monograph 143. American Geophysic al Union,

Washington, DC, pp. 49–-63. http://dx.doi.org/10.1029/143M07.

Branca, S., Coltelli, M., Groppelli, G., Lentini, F., 2011. Geological map of Etna volcano, 1:

50,000 scale. Boll. Soc. Geol. Ital. 130, 265–291.

Broquen, P., Lobartini, J .C., Candan, F., Falbo, G., 2005. Allophane, alumini um and or-

ganic matter accumulation across a bioclimatic sequence of volcanic ash soils of

Argentina. Geoderma 129, 167–177. http://dx.doi.org/10.1016/j.geoderma.2004.

12.041.

Buat-Ménard, P., Arnold, M., 1978. The heavy metal chemistry of atmospheric particulate

matter emitted by Mount Etna volcano. Geophys. Res. Lett. 5, 245–248. http://dx.doi.

org/10.1029/GL005i004p00245.

Buemi, A., Cimino, G., Strazzulla, G., 1998. Characterization of Etnean soils and application to

mars. J. Geophys. Res. Planets 103, 667–13,674. http://dx.doi.org/10.1029/98JE00816.

Busà, T., Cirrincione, R., Cristofolini, R., Pezzino, A., 1998. Caratterizzazione petrologico-

tessiturale di una successione piroclastica in prossimita di Monte Salto del Cane

(versante meridionale dell'Etna): implicazioni sui processi di alterazione e di

pedogenesi. Mineral. Petrogr. 41, 61–78.

Buurman, P., Rodeja, E.G., Cortizas, A.M., van Doesburg, J.D.J., 2004. Stratiﬁcation of parent

material in European volcanic and related soils studied by laser-diffraction grain-

sizing and c hemical analysis. Catena 56, 127 –144. http://dx.doi.org/10.1016/j.

catena.2003.10.007.

Carpenter, M.P., 2004. Early Successional Processes of Basaltic Lava Ecosystems on Mt

Etna (Sicily) With Additional Comparative Studies of Mauna Loa (Hawaii) (Ph.D. the-

sis) University of Luton.

Certini, G., Fernández-Sanjurjo, M.J., Corti, G., Ugolini, F.C., 2001. The contrasting effect of

broom and pine on pedogenic processes in volcanic soils (Mt Etna, Italy). Geoderma

102, 239–254. http://dx.doi.org/10.1016/S0016-7061(01)00017-9.

Certini, G., Wilson, M.J., Hillier, S.J., Fraser, A.R., Delbos, E., 2006. Mineral weathering in

trachydacitic-derived soils and saprolites involving formation of embryonic halloysite

and gibbsite at Mt. Amiata, Central Italy. Geoderma 133, 173–190. http://dx.doi.org/

10.1016/j.geoderma.2005.07.005.

Chester, D.K., Duncan, A.M., Guest, J.E., Kilburn, C.R.J., 1985. Mount Etna: The Anatomy of a

Volcano. Chapman and Hall, London.

Chester, D.K., Duncan, A., James, P.A., 2010. Mount Etna, Sicily: landscape evolution and

hazard responses in the pre-industrial era. In: Martini, P., Chesworth, W. (Eds.), Land-

scapes and Societies. Springer, Dordrecht.

Cimino, G., Toscano, G., 1998. Dissolution of trace metals from lava ash: inﬂuence on the

comp osition of rainwater in the Moun t Etna volcanic area. Environ. Pollut. 99,

389–393. http://dx.doi.org/10.1016/S0269-7491(98)00004-9.

Colombo, C., Sellitto, M.V., Palumbo, G., Terribile, F., Stoops, G., 2007. Characteristics and

genesis of volcanic soils from south central Italy: Mt Gauro (Phlegraean Fields, Cam-

pania) and Vico Lake (Latium). In: Arnalds, O., Bartoli, F., Buurman, P., Oskarsson, H.,

Stoops, G., García-Rodeja, E. (Eds.), Soils of Volcanic Regions in Europe. Springer, Ber-

lin, pp. 197–229.

Coltelli, M., De l Carlo, P., Vezzoli, L., 1998. Discovery of a plinian basaltic eruption of

Roman age on Etna volcano, Italy. Geology 26, 1095–1098.

Coltelli, M., Del Carlo, P., Vezzoli, L., 2000. Stratigraphic constraints for explosive activity in

the past 100 ka at Etna volcano, Italy. Int. J. Earth Sci. 89, 665–67

7. http://dx.doi.org/

10.1007/s005310000117.

Comitato per la Carta dei Suoli, 1966. Carta dei Suoli D'Italia. A. and R, Senatori, Firenze.

Condomines, M., Tanguy, J.C., Kieffer, G., Allegre, J.C., 1982. Magmatic evolution of a volca-

no studied by

230

Th-

238

U disequilibrium and trace element systematics: the Etna

case. Geochim. Cosmochim. Acta 46, 1397–1416. http://dx.doi.org/10.1016/0016-

7037(82)90275-7.

Corsaro, R.A., Pompilio, M., 2004. Dynamics of magma at Mount Etna. in: Mt Etna: Volcano

Labor atory. In: Bonaccorso, A., Calvari, S., Coltelli, M., del Negro, C., Falsaperla, S.

(Eds.), Mt Etna: Volcano Laboratory. AGU Geophys. Monogr. Ser. 143, pp. 91–110.

http://dx.doi.org/10.1029/143GM07.

Dazzi, C., 2007. Environmental features and land use of Etna (Sicily, Italy). In: Arnalds, O.,

Bartoli, F., Buurman, P., Oskarsson, H., Stoops, G., García-Rodeja, E. (Eds.), Soils of Vol-

canic Regions in Europe. Springer, Berlin, pp. 629–644.

De Beni, E., Branca, S., Coltelli, M., Groppelli, G., Wijbrans, J.R., 2011.

Ar/

Ar isotope dat-

ing of Etna volcano succession. Boll. Soc. Geol. Ital. 130, 292–305.

Dekayir, A., El Maâtaoui, M., 2000 . Apport de l'analyse d'images dans l'ét ude de

l'altération d'un basalte alcalin (Moyen Atlas, Maroc). Eng. Geol. 56, 325–334.

http://dx.doi.org/10.1016/S0013-7952(99)00099-X.

Del Carlo, P., Branca, S., 1998. Tephrostratigraphic dating of the pre-1300 AD SE ﬂank

eruption of Mt Etna. Acta Vulcanol. 10, 33–37.

Durbin, C.S., Henderson-Sellars, A., 1981. Meteorological importance of the volcanic

activity of Mount Etna. Weather 36, 284–291. http://dx.doi.org/10.1002/j.1477-8696.

1981.tb05343.x.

Egli, M., Mastrolonardo, G., Seiler, R., Raimondi, S., Favilli, F., Crimi, V., Krebs, R., Cherubini,

P., Certini, G., 2012. Charcoal and stable soil organic matter as indicators of ﬁre fre-

quency, climate and past vegetation in volcanic soils of Mt Etna, Sicily. Catena 88,

14–26. http://dx.doi.org/10.1016/j.catena.2011.08.006.

Egli, M., Mirabella, A., Alioth, L., Nater, M., Raimondi, S., 2007. Effect of climate and vege-

tation on soil organic carbon, humus fractions, allophane, imogolite, kaolinite and

oxyhydroxides in volcanic soils of Etna (Sicily). Soil Sci. 172, 673–691. http://dx.doi.

org/10.1097/ss.0b013e31809eda23.

Egli, M., Mirabella, A., Nater, M., Alioth, L., Raimondi, S., 2008. Clay minerals, oxyhydroxide

formation element leaching and humus development in volcanic soils. Geoderma

143, 101–114. http://dx.doi.org/10.1016/j.geoderma.2007.10.020.

Fernández-Sanjurjo, M., Corti, G., Certini, G., Ugolini, F., 2003. Pedogenesis induced

by Genista aetnensis (Biv.) DC on basaltic pyr oclastic deposits at different

altitudes, Mt Etna, Italy. Geoderma 115, 223–243. http://dx.doi.org/10.1016/

S0016-7061(03)00013-2.

Fierotti, G.F., Dazzi, C., Raimondi, S., 1988. Carta dei Suoli Della Sicilia, 1:250,000 (with Re-

port). Regione Siciliana Assessorato Territorio ed Ambiente and Universita Degli Studi

di Palermo, Facoltà di Agraria, Istituto di Agronomia Generale, Cattedra di Pedologia,

Palermo.

Floor, G.H., Calabrese, S., Román-Ross, G., D´Alessandro, W., Aiuppa, A., 2011. Selenium

mobilization in soils due to volcanic derived acid rain: an example from Mt Etna

volc ano, Sicily. Chem. Geol. 289, 235–244. http:// dx.doi.org/10.1016/j.chemgeo.

2011.08.004.

Francis, P., Burton, M.R., Oppenheimer, C., 1998. Remote measurements of volcanic gas

compositions by solar occultation spectroscopy. Nature 396, 567–570. http://dx.doi.

org/10.1038/25115.

García-Rodeja, E., Nóvoa, J.C., Pontevedra, X., Martínez-Cortizas, A., Buurman, P., 2007. Al-

uminium and iron fractionation of European volcanic soils by selective dissolution

techniques. In: Arnalds, O., Bartoli, F., Buurman, P., Oskarsson, H., Stoops, G., García-

Rodeja, E. (Eds.), Soils of Volcanic Regions in Europe. Springer, Berlin, pp. 325–351.

Gauthier, P.J., Le Cloarec, M.F., 1998. Variability of alkali and heavy metal ﬂuxes released

by Mt Etna volcano, Sicily, between 1991 and 1995. J. Volcanol. Geotherm. Res. 81,

311–326. http://dx.doi.org/10.1016/S0377-0273(98)00002-X.

Giammanco, S., Gurrieri, S., Principio, E., Valenza, M., 1996. CO

ﬂuxes from Mt Etna soils.

Acta Vulcanol. 8, 220–221.

Gillot, P.-Y., Kieffer, G., Romano, R., 1994. The evolution of Mount Etna in the light of po-

tassium–argon dating. Acta Volcanol. 5, 81–87.

Guest, J.E., Duncan, A.M., Stofan, E.R., Anderson, S.W., 2012. Effect of slope on develop-

ment of pahoehoe ﬂow ﬁelds: evidence from Mount Etna. J. Volcanol. Geotherm.

Res. 219-220, 52–62. http://dx.doi.org/10.1016/j.jvolgeores.2012.01.006.

Hamidi, E.M., Nahon, D., McKenzie, J.A., Micard, A., Colin, F., Kamel, S., 1999. Marine Sr

(Ca) input in quaternary volcanic rock weathering proﬁles from the Mediterranean

coast of Morocco: Sr isotopic approach. Terra Nova 11, 157–161. http://dx.doi.org/

10.1046/j.1365-3121.1999.00239.x.

Heiken, G., Wohletz, K., 1985. Volcanic Ash. University of California Press, Berkeley.

Hesse, P.R., 1971. A Textbook of Soil Chemical Analysis. Murray, London.

Hopkins, D.W., Bartoli, F., 2004. Size and activity of the soil microbial community from a

range of European volcanic soils. In: Oskarsson, H., Arnalds, O. (Eds.), Volcanic Re-

sources in Europe. Cost Action 622 Final Meeting. Abstracts. Agricultural Research In-

stitute, Rejkyavic, pp. 35–36 (http://www.rala.is/Andosol/Fjolrit214.pdf Accessed 31

July, 2015).

Hopkins, D.W., Badalucco, L., English, L.C., Meli, S.M., Chudek, J.A., Ioppolo, A., 2007. Plant

litter decomposition and microbial characteristics in volcanic soils (Mt Etna, Sicily) at

different stages of development. Biol. Fertil. Soils 43, 461–469. http://dx.doi.org/10.

1007/s00374-006-0124-3.

Huggett, R.J., 1998. Soil chronosequences, soil development, and soil evolution: a critical

review. Catena 32, 152–172. http://dx.doi.org/10.1016/S0341-8162(98)00053-8.

IUSS Working Group WRB, 2006. World Reference Base for Soil Resources 2006. World

Soil Resources Reports No. 103. FAO, Rome. (ftp://ftp.fao.org/agl/agll/docs/

wsrr103e.pdf Accessed 31 July, 2015).

James, P.A., Atherton, M.A., Harvey, A.M., Firmin, A., Morrow, A., 1997. In: M ee, C.B.,

Forbes, H.A. (Eds.), A Rough and Rocky Place: The Landscape and Settlement History

of the Methana Peninsula, Greece. University of Liverpool Press, Liverpool, pp. 5–32.

Johnson, D.L., 1985. Soil thickness processes. Catena Suppl. 6, 29–40.

Kilburn, C.R.J., Guest, J.E., 1993. Aa lavas of Mount Etna. In: Kilburn, C.R.J., Luongo, G.

(Eds.), Active Lavas: Monitoring and Modelling. UCL Press, London, pp. 73–103.

Legros, J.-P., 2013. Major Soil Groups of the World. Ecology, Genesis, Properties and Clas-

siﬁcation. CRC Press, Taylor and Francis, Boca Raton, Florida.

Leighton, R., 1999. Sicily Before History: An Archaeological Survey From the Palaeolithic

to the Iron Age. Duckworth, London.

Leonardi, S., Rapp, M., 1981. Return of mineral elements and nitrogen to the soil through

leav es in an ecosystem of Quercus ilex L. on Mount-Mi nardo (Etna). Flora 171,

329–337.

Leonardi, S., Rapp, M., Denes, A., 1992. Organic-matter distribution and ﬂuxes within a

holm oak (Quercus ilex

L.) stand in the Etna Volcano — a

synthesis. Vegetatio 100,

219–224. http://dx.doi.org/10.1007/BF00118228.

Leonardo, S., Rapp, M., Failla, M., Guarnaccia, D., 1996. Biomass productivity and organic

matter ﬂuxes in Castanea sativa Mill stands of various elevations in the Etna volcano.

Ann. Sci. For. 53, 1031–1048. http://dx.doi.org/10.1051/forest:19960510.

Liuzzo, M., Gurrieri, S., Giudice, G., Giuffrida, G., 2013. Ten years of soil CO

continuous

monitoring on Mt Etna; exploring the relationship between processes of soil

degassing and volcanic activity. Geochem. Geophys., Geosyst. G. 14 (super 3),

2886–2899. http://dx.doi.org/10.1002/ggge.20196.

Lo Papa, G., Palermo, V., Parisi, S., Laudacina, V.A., Tusa, D., Scalenghe, R., 2003.

Caratteristiche di una sequenza di suoli forestali nel versante Nord-occidentale

dell'Etna. Bol. Soc. Ital. Sci. Suolo 52, 513–522.

Lulli, L., 2007. Italian volcanic soils. In: Arnalds, O., Bartoli, F., Buurman, P., Oskarsson, H.,

Stoops, G., García-Rodeja, E. (Eds.), Soils of Volcanic Regions in Europe. Springer, Ber-

lin, pp. 51–67.

Malucelli, F., Terribile, F., Colombo, C., 1999. Mineralogy, micromorphology and chemical

analysis of andosols on the Island of Sao Mi guel (Azores). Geoderma 88, 73–98.

http://dx.doi.org/10.1016/S0016-7061(98)00081-0.

Martínez Cortizas, A., García-Rodeja, E., Gayosoa, E., Nóvoa Muñoz, J.C., Po ntevedra

Pombal,X.,Buurman,P.,Terribile,F.,2003. Distribution of some selected major

and trace elements in four I talian soils dev eloped from the deposits of the

Gauro and Vico volcanoes. Geoderma 117, 215–224. http://dx.doi.org/10.1016/

S0016-7061(03)00124-1.

Martínez C ortizas, A., Nóvoa, J.C., Pontevedra, X., T aboada, T., G arcía-Rodreja, E.,

Chesworth, W., 2007. Elemental composition of reference European volcanic soils.

296 P. James et al. / Catena 137 (2016) 277–297

In: Arnalds, O., Bartoli, F., Buurman, P., Oskarsson, H., Stoops, G., García-Rodeja, E.

(Eds.), Soils of Volcanic Regions in Europe. Springer, Berlin, pp. 289–306.

Mastrolonardo, G., Certini, G., Krebs, R., Forte, C., Egli, M., 2013. Effects of ﬁre on soil organic

matter quality along an altitudinal sequence on Mt Etna, Sicily. Catena 110, 133–145.

McKeague, J.A., 1967. An evaluation of 0.1 M pyrophosphate and pyrophosphate–dithi onite

in comparison with oxalate as extractants of the accumulation products in podzols and

someothersoils.Can.J.SoilSci.47,95–99. http://dx. doi.org/10.4141/cjss67-017.

McKeague, J.A., Day, J.H., 1966. Dithionite- and oxalate-extractable Fe and Al as aids in dif-

ferentiating various classes of soils. Can. J. Soil Sci. 46, 13–22. http://dx.doi.org/10.

4141/cjss66-003.

Mehra, O.P., Jackson, M.L., 1960. Iron oxide removal from soils and clays by a dithionite–

citrate system buffered with sodium carbonate. Clay Clay Miner. 7, 317–327.

Mizota, C., van Reeuwijk, L.P., 1989. Clay Mineralogy and Chemistry of Soils Formed in

Volcanic Material in Diverse Climatic RegionsSoil Monograph 2. International Soil

Reference and Information Centre, Wageningen (185 pp).

Nanzyo, M., Shoji, S., Dahlgren, R., 1993. Physical characteristics of volcanic ash soils. In:

Shoji, S., Nanzyo, M., Dahlgren, R.A. (Eds.), Volcanic Ash Soils: Genesis, Properties

and Utilization. Elsevier, Amsterdam, pp. 189–207.

Palumbo, B., Angelone, M., Bellanca, A., Dazzi, C., Hauser, S., Neri, R., Wilson, J., 2000. Inﬂu-

ence of inheritance and pedogenesis on heavy metal distribution in soils of Sicily,

Italy. Geoderma 95, 247–266. http://dx.doi.org/10.1016/S0016-7061(99)00090-7.

Parisi, S., Meijer, E., Dazzi, C., Buurman, P., 2003. Una Metodologia Auto Validante per la

Quantiﬁcazione Della Frazione Amorfa Inorganica in Suoli Dell'Etna. Paper Presented

at the Annual Convention of the Società Italiana Della Scienza del Suolo, ‘Qualità del

Suolo, Impatto Antropico e Qualità dei Prodotti Agricoli’, Siena, June 2003. Istituto

Sperimentale per lo Studio e la Difesa del Suolo and the Amministrazione Provinciale

di Siena.

Polacci, M., Papale, P., 1999. The development of compound lava ﬁelds at Mount Etna.

Phys. Chem. Earth Solid Earth Geod. 24, 949–952. http://dx.doi.org/10.1016/S1464-

1895(99)00140-4.

Poli Marchese, E., Patti, G., 2000. Car ta della vegetazione dell’ Etna a scala 1:50,000.

Unversità di Catania, Selca, Firenze.

Puzzolo, V., Folving, S., 2001. A GIS integration of topographic, geological and climatic data

for the analysis of the Mt Etna forests distribution (Sicily, Italy). (http://www.ec-gis.

org/Workshops/7ec-gis/papers/html/puzzolo/virginia.htm Accessed 31 July, 2015).

Quantin, P., Lulli, L., Dabin, B., Bidini, D., Bouleau, A., 1985. Characteristics and genesis of

two andosols in central Italy. Catena Suppl. 7, 107–117.

Romano, R., Guest, J.E., 1979. Volcanic geology of the summit and northern ﬂank of Mount

Etna, Sicily. Boll. Soc. Geol. Ital. 98, 189–215.

Sadori, L., Narcisi, B., 2001. The postglacial record of environmental history from Lago di

Pergusa, Sicily. The Holocene 11, 655–671.

http://dx.doi.org/10.1191/09596830195681.

nta Regina, I., Leonardi, S., Rapp, M., 2001. Foliar nutrient dynamics and nutrient-use ef-

ﬁciency in Castanea sativa coppice stands of southern Eu rope. Forestry 74, 1–10.

http://dx.doi.org/10.1093/forestry/74.1.1.

Schoeneberger, P.J., Wysocki, D.A., Benham, E.C., Soil Survey Staff, 2012. Field Book for De-

scribing and Sampling Soils, Version 3.0. Natural Resources Conservation Service, Na-

tional Soil Survey Center, Lincoln, NE (http://www.nrcs.usda.gov/wps/portal/nrcs/

detail/soils/ref/?cid=nrcs142p2_054184. Accessed 27 July, 2015).

Schwertmann, U., 1985. The effect of pedogenic environments on iron oxide minerals. In:

Stewart, B.A. (Ed.)Advances in Soil Sci. 1. Springer Verlag, New York, pp. 171–200.

http://dx.doi.org/10.1007/978-1-4612-5046-3_5

Shillam, L., Hopkins, D.W., Badalucco, L., Laudicina, V.A., 2008. Structural diversity and en-

zyme activity of volcanic soils at different stages of development and response to

experimental disturbance. Soil Biol. Biochem. 40, 2182–2185. http://dx.doi.org/10.

1016/j.soilbio.2008.04.006.

Shoji, S., Nanzyo, M., Dahlgren, R., 1993. Volcanic Ash Soils: Genesis, Properties and Utili-

zation. Elsevier Science, Amsterdam.

Singer, A., Zarei, M., Lange, F.M., Stahr, K., 2004. Halloysite characteristics and formation in

the northern Golan Heights. Geoderma 123, 279–295. http://dx.doi.org/10.1016/j.

geoderma.2004.02.012.

Soil Survey Staff, 1975. Soil Taxonomy. U.S.D.A., Soil Conservation Service, Agriculture

Handbook, 436, U.S. Government Printing Ofﬁce, Washington, D.C.

Soil Survey Staff, 1999. Soil Taxonomy: A Basic System of Soil Classiﬁcation for Making

and Interpreting Soil Surveys. 2nd Edition. Natural Resources Conservation Service.

U.S. Department of Agriculture Handbook 436. (http://www.nrcs.usda.gov/wps/

portal/nrcs/main/soils/survey/class/taxonomy/ Accessed 31 July, 2015.).

Soil Survey Staff, 2014. Keys to Soil Taxonomy. 12th edition. United States Department of

Agriculture, Natural Resources Conservation Service, Washington, DC (http://www.

nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/class/?cid=n rcs142p2_053580

Accessed 31 July, 2015).

Taboada, T., García, C., Martínez-Cortizas, A., Nóvoa, J.C., Pontevedra, X., García-Rodeja, E.,

2007. Chemical weathering of reference European volcanic soils. In: Arnalds, O.,

Bartoli, F., Buurman, P., Oskarsson, H., Stoops, G., García-Rodeja, E. (Eds.), Soils of Vol-

canic Regions in Europe. Springer, Berlin, pp. 307–323.

Tanguy, J.C., Condomines, M., Branca, S., La Delfa, S., Coltelli, M., 2012. New

archeomagnetic and

226

Ra–

230

Th dating of recent lavas for the geological map of

Etna volcano. Ital. J. Geosci. 131 (2), 241–257. http://dx.doi.org/10.3301/IJG.2012.01.

Tanguy, J.C., Condomines, M., Kieffer, G., 1997. Evolution of the Mount Etna magma: con-

straints on the present feeding system and eruptive mechanism. J. Volcanol.

Geotherm. Res. 75, 221–250. http://dx.doi.org/10.1016/S0377-0273(96)00065–0.

Tanguy, J.C., Condomines, M., Le Goff, M., Chillemi, V., La Delfa, S., Patanè, G., 2007. Mount

Etna erupti ons of the last 2750 years: revised chronology and location through

archeomagnetic and

226

Ra–

230

Th dating. Bull. Volcanol. 70, 55–83. http://dx.doi.org/

10.1007/s00445-007-0121-x.

Vacca, A., Adamo, P., Pigna, M., Violante, P., 2003. Genesis of tephra-derived soils from the

Roccamonﬁna Volcano, south central Italy. Soil Sci. Soc. Am. J. 67, 198–207. http://dx.

doi.org/10.2136/sssaj2003.0198.

Varekamp, J.C., Thomas, E., Germani, M., Buseck, P.R., 1986. Particle geochemistry of vol-

canic plumes of Etna and Mount St Helens. J. Geophys. Res. Solid Earth 91,

12223–12248. http://dx.doi.org/10.1029/JB091iB12p1223.

Varrica, D., Aiuppa, A., Dongarra, G., 2000. Volcanic and anthropogenic contribution to

heavy metal content in lichens from Mt Etna and Vulcano island (Sicily). Environ.

Pollut. 108, 153–162. http://dx.doi.org/10.1016/S0269-7491(99)00246–8.

Vingiani, S., Terribile, F., Adamo, P., 2013. Weathering and particle entrapm ent at the

rock–lichen interface in Italian volcanic environments. Geoderma 2 07, 244–255.

http://dx.doi.org/10.1016/j.geoderma.2013.05.015.

Vitousek, P.M., Chadwick, O.A., Crews, T., Fownes, J., Hendricks, D., Herbert, D., 1997. Soil

and ecosystem development across the Hawaiian Islands. GSA Today 7, 1–8(ftp://

rock.geosociety.org/pub/GSAToday/gt9709.pdf Accessed 31 July, 2015).

Yaalon, D.H., 1997. Soils in the Mediterranean region: what makes them different? Catena

28, 157–169. http://dx.doi.org/10.1016/S0341-8162(96)00035–5.

Zehetner, F., Miller, W.P., West, L.T., 2003. Pedogenesis of volcanic ash soils in Andean

Ecuador. Soil Sci. Soc. Am. J. 67, 1797–1809. http://dx.doi.org/10.2136/sssaj2003.

1797.

297P. James et al. / Catena 137 (2016) 277–297